Machine Learning Big Data

Agarwood oil is consumed during traditional ceremonies and even in medicinal purposes due to its effective therapeutic characteristic. As a part of ongoing research on agarwood oil, this paper presented a k-nearest neighbor (k-NN) modelling of agarwood oil samples available in the capital of Malaysia market. The work involved agarwood oil samples from three sources which are lab, local manufacturer and market. The inputs are the chemical compounds and the output is the oil's resources. The input-output was divided into training and testing dataset with the ratio of 80% to 20%, respectively, before they were fed to the k-NN for model development as well as model validation. During the model development, the k-value was varied from 1 to 5, and their accuracy was observed. The result showed that the k=1 and k=2 shared the similar accuracy for training and testing datasets, which are 98.63% and 100.00%, respectively. This study revealed the capabilities of the k-NN model in classifying the agarwood oil samples to the three sources: lab, local manufacturer and market. It was a significant study and contributed to further work especially those related to agarwood oil research area.

Risk modelling along with multi-objective optimization problems have been at the epicenter of attention for supply chain managers. In this paper, we introduce a dataset for risk modelling in sophisticated supply chain networks based on formal mathematical models. We have discussed the methodology and simulation tools used to synthesize the dataset. Additionally, the underlying mathematical models are discussed in granular details along with providing directions to conducting statistical analyses or neural machine learning models. The simulation is performed using MATLAB ™Simulink and the models are illustrated as well.

The rapid increase in the size of databases required to store power quality monitoring data has demanded new techniques for analysing and understanding the data. One suggested technique to assist in analysis is data mining. Data mining is a process that uses a variety of data analysis tools to identify hidden patterns and relationships within large samples of data. This paper presents several data mining tools and techniques that are applicable to power quality data analysis to enable efficient reporting of disturbance indices and identify network problems through pattern recognition. This paper also presents results of data mining techniques applied to power quality data from an MV electrical distribution system to identify disturbance patterns.

While advanced analysis of large dataset is in high demand, data sizes have surpassed capabilities of conventional software and hardware. Hadoop framework distributes large datasets over multiple commodity servers and performs parallel computations. We discuss the I/O bottlenecks of Hadoop framework and propose methods for enhancing I/O performance. A proven approach is to cache data to maximize memory-locality of all map tasks. We introduce an approach to optimize I/O, the in-node combining design which extends the traditional combiner to a node level. The in-node combiner reduces the total number of intermediate results and curtail network traffic between mappers and reducers.

Effective communication occurs when the receiver and sender both understand and synchronize the flow of information across board. The utility of language extends beyond human to human interaction and includes also, the use of syntactically formed programming languages to interact with digital systems. Nigeria has an estimate of over 450 languages, which makes it cumbersome to harmonize and put all into a single large repository for data mining. The goal of this paper is to firmly establish the importance of Information Technology in galvanizing Nigerian Languages and Mining scientific data thereof. The purpose of applying Information and Communication Technology (ICT) is to codify the process of extracting various underlying meanings in a language, processing the various idioms, proverbs and quaint statements in such language with the view of bringing out the creativity behind them. The authors explore the developmental stages and techniques of applying an artificial Intelligence system that scans through a given indigenous linguistic system to bring out the hidden facts therein. It is recommended that stakeholders in the 'digital humanities' adopt such mining platforms which helps in achieving greater insight into the diverse cultures and languages, in turn, promoting easy learning experience for indigenous languages.

Il 5 dicembre 2017 il robot AlphaZero ha vinto ripetutamente  contro Stockfish, il (fino a quel momento) migliore giocatore automatico di scacchi - più precisamente, il Top Chess Engine Championship del 2016. AlphaZero è un algoritmo di deep-learning: invece di valutare tutte le possibilità del tavolo per decidere la mossa migliore, il robot parte da una tabula rasa, su cui vengono inserite solo le regole fondamentali del gioco, dato che AlphaZero si auto-insegna la strategia affrontando innumerevoli partite contro sé stesso. Ci sono volute solo ventiquattro ore perché il robot fosse in grado di giocare a un livello «sovrumano» e superasse ogni giocatore mai esistito, umano o computerizzato. I creatori di AlphaZero sono ricercatori presso DeepMind, azienda gestita da Alphabet Inc., a sua volta società madre di Google. Anche Google usa degli algoritmi di apprendimento, che si nutrono delle informazioni fornite dagli utenti, le quali vanno a formare i cosiddetti Big Data, vale a dire enormi quantità di dati all'interno dei quali vengono identificati dei trend in grado di spiegare «come agiremo di qui a poco». Come AlphaZero diventa un giocatore migliore dopo ogni partita giocata contro sé stesso, Google (ma anche le altre «techno-corporation») crea un profilo sempre più preciso di ogni suo utente ogni volta che tale utente fa uso di un suo servizio (dal motore di ricerca, alle mappe, all'indirizzo email). Il modo in cui techno-corporation come Google, Facebook, Amazon, Airbnb e «altre decine di piattaforme che utilizziamo sempre» raccolgono e usano i nostri dati è un tema centrale nel Manuale di Nicola Zamperini, ma più centrale ancora è la riflessione sul bassissimo livello di consapevolezza che l'utente medio ha di questo processo. Il volume di Zamperini è una chiamata alla presa di coscienza in primis riguardo al fatto che nelle transazioni quotidiane con le techno-corporation noi da una parte riceviamo servizi, ma dall’altra siamo anche la «materia prima». La disobbedienza, o meglio l’imprevedibilità, viene presentata come caratteristica fondamentalmente umana e incomprensibile a un computer. Zamperini propone di sfruttarla a nostro vantaggio per confondere gli algoritmi di apprendimento: incita alla «diserzione», a inserire di proposito l’errore, e suggerisce di «praticare gesti di follia a casaccio, momenti di imprevedibilità digitale e inversioni di rotta non programmate».

Dynamically changing background ("dynamic background") still presents a great challenge to many motion-based video surveillance systems. In the context of event detection, it is a major source of false alarms. There is a strong need from the security industry either to detect and suppress these false alarms, or dampen the effects of background changes, so as to increase the sensitivity to meaningful events of interest. In this paper, we restrict our focus to one of the most common causes of dynamic background changes: that of swaying tree branches and their shadows under windy conditions. Considering the ultimate goal in a video analytics pipeline, we formulate a new dynamic background detection problem as a signal processing alternative to the previously described but unreliable computer vision-based approaches. Within this new framework, we directly reduce the number of false alarms by testing if the detected events are due to characteristic background motions. In addition, we introduce a new dataset suitable for the evaluation of dynamic background detection. It consists of real-world events detected by a commercial surveillance system from two static surveillance cameras. The research question we address is whether dynamic background can be detected reliably and efficiently using simple motion features and in the presence of similar but meaningful events such as loitering. Inspired by the tree aerodynamics theory, we propose a novel method named local variation persistence (LVP), that captures the key characteristics of swaying motions. The method is posed as a convex optimization problem whose variable is the local variation. We derive a computationally efficient algorithm for solving the optimization problem, the solution of which is then used to form a powerful detection statistic. On our newly collected dataset, we demonstrate that the proposed LVP achieves excellent detection results and outperforms the best alternative adapted from existing art in the dynamic background literature.

This project examines selected big data technologies from an enterprise usage viewpoint.
The addressed use cases include data warehousing, visualization, reporting and integration.
Industry trending big data technologies are examined in terms of their ability to address the needs of organizations.
The emphasis of this project is on the practical usage of big data technologies in analysing business data and thus achieving results that enhance business value for organizations. Said results could be a factor of business insight gained or competitive advantage gained from using newer and better technologies.

Therefore, this project will focus on the steps involved in setting up realistic use cases and the resultant outputs.
Firstly an overview of big data technologies is given with historical references followed by steps involved in setting up development and runtime environments.
Next the use cases are developed one by one and the results are used to justify the need for big data technologies in organizations.
Finally a use case demonstrating how business data stored in big data repositories could be shared within and outside of an organization using industry standards like REST, SOAP etc.



The end results are as follows :
(1) An appreciation of how big data technologies could be efficiently used in storing and analysing data for the purpose of enhancing business value for organizations,
(2) A demonstration of possible use cases showing how big data technologies could be employed in analysing business data,
(3) Enabling data scientists in organizations to take advantage of the latest technologies in analysing data efficiently.

As an occurrence that jeopardises vital national interests or the basic needs of the populace, a crisis necessitates rapid decision-making and coordination between various departments and agencies in order to resolve it effectively. As a result, crisis and disaster management systems are necessary and critical. Crisis and disaster response systems are intricate, requiring numerous phases, techniques, and resources. These systems require useful and necessary data that can be used to make future decisions more effectively, such as historical and current data on crises. The use of machine learning and big data technologies to process data from crises and disasters has the potential to yield significant results in this area. The first section of this document discusses crisis management systems and available tools, such as big data and machine learning. Additionally, a machine learning and big data approach to crisis management systems were developed, which included a description and experiments, as well as a discussion of the findings and the field's future directions.

Data  communications  over  the  Internet  face serious  threats  to  countries  with  repressive  regimes,  leading them  to  develop  and  deploy  censorship  mechanisms  within their  networks.  Unfortunately,  existing  censorship  systems  do not  provide  high  availability  guarantee  to  their  users,  as censors can easily identify, hence disrupt, the traffic belonging to  these  systems  using  today’s  advanced  censorship technologies.  In  this  paper,  we  propose  serving  the  Web  by Exploiting  Email  Tunnels  (SWEET),  a  highly  available censorship-resistant    infrastructure.    SWEET    works    by encapsulating a censored user’s traffic inside email messages that  are  carried  over  public  email  services  like  Gmail  and Yahoo Mail.  As the operation of SWEET is not bound to any specific  email  provider,  we  argue  that  a  censor  will  need  to block  email  communications  all  together  in  order  to  disrupt SWEET,  which  is  unlikely  as  email  constitutes  an  important part of today’s Internet. Through experiments with a prototype of  our  system,  we  find  that  SWEET’s  performance  is sufficient  for  Web  browsing.  In  particular,  regular  Websites are downloaded within couple of seconds.

Risk modelling along with multi-objective optimization problems have been at the epicenter of attention for supply chain managers. In this paper, we introduce a dataset for risk modelling in sophisticated supply chain networks based on formal mathematical models. We have discussed the methodology and simulation tools used to synthesize the dataset. Additionally, the underlying mathematical models are discussed in granular details along with providing directions to conducting statistical analyses or neural machine learning models. The simulation is performed using MATLAB ™Simulink and the models are illustrated as well.

In today's world, due to the advancement of technology, predicting the students' performance is among the most beneficial and essential research topics. Data Mining is extremely helpful in the field of education, especially for analyzing students' performance. It is a fact that predicting the students' performance has become a severe challenge because of the imbalanced datasets in this field, and there is not any comparison among different resampling methods. This paper attempts to compare various resampling techniques such as Borderline SMOTE, Random Over Sampler, SMOTE, SMOTE-ENN, SVM-SMOTE, and SMOTE-Tomek to handle the imbalanced data problem while predicting students' performance using two different datasets. Moreover, the difference between multiclass and binary classification, and structures of the features are examined. To be able to check the performance of the resampling methods better in solving the imbalanced problem, this paper uses various machine learning classifiers including Random Forest, K-Nearest-Neighbor, Artificial Neural Network, XG-boost, Support Vector Machine (Radial Basis Function), Decision Tree, Logistic Regression, and Naïve Bayes. Furthermore, the Random hold-out and Shuffle 5-fold cross-validation methods are used as model validation techniques. The achieved results using different evaluation metrics indicate that fewer numbers of classes and nominal features will lead models to better performance. Also, classifiers do not perform well with imbalanced data, so solving this problem is necessary. The performance of classifiers is improved using balanced datasets. Additionally, the results of the Friedman test, which is a statistical significance test, confirm that the SVM-SMOTE is more efficient than the other resampling methods. Moreover, The Random Forest classifier has achieved the best result among all other models while using SVM-SMOTE as a resampling method. INDEX TERMS Classification, data mining, educational data mining, imbalanced data problem, machine learning, resampling methods, statistical analysis.

This comprehensive edited advanced book will cover machine and deep learning strategies for security enhancement in wireless Sensor Network. This topic has matured with very exciting developments and applications in real time including Ubiquitous & Pervasive Computing; Wireless Systems; Computer Science and Information Technology; Computer Engineering; Networking and Telecommunications; Mobile and Wireless Computing; Security and Forensics; Cyber and Network Security; IT Security and Ethics; etc. The goal of this new volume is to summarize the recent research advances in the field. We believe now is the time for an advanced book that will cover the state of the art in the field. Our aim is to edit a useful refer Strategies for Security Enhancement in Wireless Sensor Network for a cross disciplinary audience working in Computer Vision, and Machine and Deep Learning research in both industry and academia. Each chapter is expected to be self-contained.

Health is considered to be a fundamental necessity of all living beings, particularly humans, hence its availability is a critical component of the Human Development Index (HDI) measurement of human health. Neonatal (NMR) and infant (IMR) mortality rates are not subjects that take too much time or attention for most of us on our day-today thinking, but they are also vital to guaranteeing our quality of living. Evidence-based awareness of infant mortality patterns and drivers can support correct strategies required to counter the threat. Nigeria's attempts to minimize under-five mortality have been skewed against neglect of neonates in favor of childhood mortality, and as such the literature lacks adequate knowledge to estimate neonatal rates accurately. Whereas studies have shown that in the neonatal period, about half of infant deaths occur. Awareness of the rate of neonatal deaths as well as neonatal mortality determinants are important for the design of intervention programs to improve neonatal survival. Hence, this research was conducted to apply Random Forest Regression in predicting neonatal death rates to help better prepare healthcare systems for future occurrences. This model will be educated using the existing dataset of actual neonatal death rates from 1970 through to 2018. This model may be used to predict neonatal death rates and thus help prepare them adequately to reduce risk by providing medical treatment, thus reducing infant mortality levels.

در عصر فناوری ‌اطلاعات و ارتباطات، نگرش مصرف‌کنندگان به ادراک از خلاقیت پیامک‌های تبلیغاتی، یکی از موضوعات مطرح در حوزه بازاریابی است و هنوز شناخت چندانی درباره آن وجود ندارد. هدف این پژوهش، بررسی تأثیر ادراک از خلاقیت پیامک‌های تبلیغاتی بر نگرش و واکنش مصرف‌کنندگان به پیامک‌های تبلیغاتی ارسالی به تلفن همراه آنهاست. جامعه آماری این تحقیق کاربران تلفن همراه در شهر تهران و حجم نمونه ۳۸۵ نفر است که به‌صورت نمونه‌گیری دردسترس انتخاب شدند. تحقیق حاضر از نظر هدف، کاربردی و از نظر نحوه گردآوری‌ داده‌ها توصیفی- پیمایشی است. مدل مورد مطالعه با استفاده از مدل‌سازی معادلات ساختاری و رویکرد کمترین مربعات جزئی آزمون شد. نتایج بیانگر آن است که ادراک از خلاقیت پیامک تبلیغاتی بر نگرش و واکنش‌رفتاری مصرف‌کنندگان به پیامک‌های تبلیغاتی ارسالی به تلفن همراه آنان تأثیر مثبت معناداری دارد، درحالی‌که ادراک از خلاقیت پیامک تبلیغاتی به‌صورت غیرمستقیم و از طریق متغیر میانجی نگرش نیز بر واکنش ‌رفتاری مصرف‌کنندگان تأثیر می‌گذارد.

Our aim is to estimate the perspective-effected geometric distortion of a scene from a video feed. In contrast to most related previous work, in this task we are constrained to using low-level, spatio-temporally local motion features only. This particular challenge arises in many semi-automatic surveillance systems which alert a human operator to potential abnormalities in the scene. Low-level, spatio-temporally local motion features are sparse (and thus require comparatively little storage space) and sufficiently powerful in the context of video abnormality detection to reduce the need for human intervention by more than 100-fold. This paper introduces three significant contributions: (i) we describe a dense algorithm for perspective estimation which uses motion features to estimate the perspective distortion at each image locus and then polls all such local estimates to arrive at the globally best estimate, (ii) we also present an alternative coarse algorithm which subdivides the image frame into blocks, and uses motion features to derive block-specific motion characteristics and constrain the relationships between these characteristics, with the perspective estimate emerging as a result of a global optimization scheme, and (iii) we report the results of an evaluation using nine large sets acquired using existing close-circuit television (CCTV) cameras, not installed specifically for the purposes of this work. Our findings demonstrate that both of the proposed methods are successful, their accuracy matching that of human labelling using complete visual data (by the constraints of the setup unavailable to our algorithms).

Logistic regression analysis of factors contributing to bank deposit subscription and direct marketing success.

View project at:

https://github.com/JackKRasmus-Vorrath/Logistic_Regression_Direct-Marketing_Bank-Term-Deposit-Subscription

Nowadays, the mobile phone is one of the fastest growing technologies in the developing world with global penetration rates reaching 90% [1]. This makes it a huge warehouse for customer's data. That is, every action taken by the customer (short message service (SMS), Call or Internet session) gets recorded within the telecom operator, in the so called (CDRs). There are many types of CDRs used mainly by telecom billing systems. CDR contains a lot of information, (type of event, who is involved in this event, datetime, cell identifier where this event has taken place). This raw data represents a valuable source for analyzing human and social behavior [2]. In the agricultural domain [3] mobile phone data is used to analyze mobility and seasonal activity patterns related to livelihood zones in Senegal, by creating mobility profiles for population and segmentation. While in energy domain [4] this data is used to analyze human activity, facilitate population growth estimation in rural areas and extrapolate electricity needs. In health sector [5, 6] mobile phone data is used to study the relation between human mobility

Ethereum smart contracts have recently received new commercial applications and a lot of attention from the scientific community. Ethereum eliminates the requirement for a trusted third party by allowing untrusted parties to expose contract details in computer code. Nonetheless, as online commerce grows, plenty of fraudulent activities, such as money laundering, bribery, and phishing, emerge as major threats to trade security. For correctly recognizing fraudulent transactions, this paper developed a Light Gradient Boosting Machine (LGBM) technique-based model. The modified LGBM model optimized the parameters of Light GBM using the Euclidean distant structured estimation approach. This paper also examines the performance of different popular models such as Random Forest (RF), Multi-Layer Perceptron (MLP), Logistic Regression, k-Nearest Neighbors (KNN), XGBoost, Support Vector Classification (SVC), and ADAboost with limited features and compares their performance metrics with the proposed model for Ethereum fraudulent activity classification. A comparative performance evaluation matrices scores of different popular models along with the proposed model demonstrated the applicability of the proposed approach. The modified LGBM algorithms and RF models demonstrate the best performance compared to other models with the highest accuracies, while the modified LGBM algorithm has a slightly superior performance of 99.17 percent compared to the RF model's 98.26 percent.

Machine Learning and Applications: An International Journal (MLAIJ) is a quarterly open access peer-reviewed journal that publishes articles which contribute new results in all areas of the machine learning. The journal is devoted to the publication of high quality papers on theoretical and practical aspects of machine learning and applications.The goal of this journal is to bring together researchers and practitioners from academia and industry to focus on machine learning advancements, and establishing new collaborations in these areas. Original research papers, state-of-the-art reviews are invited for publication in all areas of machine learning.
Authors are solicited to contribute to the journal by submitting articles that illustrate research results, projects, surveying works and industrial experiences that describe significant advances in the areas of machine learning.

Ce document, constitue un support pédagogique pour étudiants et formateurs en analyse des données avec le logiciel SPSS version 26.
Ce module couvre des théories statistiques ainsi que des procédures pratiques pour encoder, traiter les données et réaliser des analyse statistiques appropriés avec le logiciel SPSS.

Automatic Student performance prediction is a crucial job due to the large volume of data in educational databases. This job is being addressed by educational data mining (EDM). EDM develop methods for discovering data that is derived from educational environment. These methods are used for understanding student and their learning environment. The educational institutions are often curious that how many students will be pass/fail for necessary arrangements. In previous studies, it has been observed that many researchers have intension on the selection of appropriate algorithm for just classification and ignores the solutions of the problems which comes during data mining phases such as data high dimensionality, class imbalance and classification error etc. Such types of problems reduced the accuracy of the model. Several well-known classification algorithms are applied in this domain existing models a student performance prediction model based on supervised learning decision tree classifier. In addition, an ensemble method is applied to improve the performance of the classifier. Ensemble methods approach is designed to solve classification, prediction problem. we propose a method for predicting final grades of students by a Recurrent Neural Network (RNN) from the log data stored in the educational systems. We applied this method to the log data from 108 students and examined the accuracy of prediction. From the experimental results, comparing with multiple regression analysis, it is confirmed that an RNN is effective to early prediction of final and suitable job for the student based on their academic performance and knowledge on skill set.

The method of instruction used by the teacher can potentially motivate or dispirit learners' interests in
a course and could also influence their learning outcomes. The traditional teaching approach is
beginning to prove ineffective as seen during the recent COVID-19 pandemic, hence the need for adoption of more flexible and interactive approach like the inquiry-based approach. The study
examines the impact of inquiry-based teaching on Computer Science learning. A total of 50 Second year students taking Computer Application Course (COS 234) in Coal City University participated in the study. They were selected using simple random sampling technique, and were divided into two major groups; Experimental Group (EG) and Control group (EG). The EG was taught using the
inquiry-based teaching approach, while the CG was taught using conventional approach. We conducted an Achievement test to investigate the effects of Inquiry-based approach on the students' learning outcomes. The post-test results show that inquiry-based teaching had significant positive
effects on the students' learning interests and academic achievements. The post-test mean score of the experimental group indicates higher academic achievements than that of the control group, showing
an increase in both students' learning interests and academic achievement after the treatment. The
study establishes that inquiry-based teaching approach presents a transformative alternative to the prescriptive didactic method by giving more opportunities to learners to engage in self-directed
learning, exploration and discovery of many possible ways to solve problems on their own.

The entire human activity is generating about 2.5 quintillion bytes of data per exponentially. The tremendous volume of data is getting generated by different banking services, including real estate loan, deposit accounts services, credit card issuing service, commercial loan lending and many more. The bank data is used for business analytic that keeps eyes on past growth also Predictive analysis opens doors for future growth. For this project, during the predictive analysis, we used some historical data generated by an organization during loan processing. We intend not. The analysis was done by building Machine Learning Random Forest(RF), Logistic Regression(LR), Support Vector Machine(SVM), Dec XG Boost. The classifier models were trained and tested on hyper analyzed using Confusion Matrix. We compared efficiencies using Recall, Precision, Accuracy and F1 score in a tabular form among all models. LR yields an F1 score of 0.90 and 0.62 for class 1 (approved) and class 0 (not Approved), respectively. Also, XG

Crime analysis and prediction is a systematic approach for identifying the crime. This system can predict region which have high probability for crime occurrences and visualize crime prone area. Using the concept of data mining we can extract previously unknown, useful information from an unstructured data. The extraction of new information is predicted using the existing datasets. Crimes are treacherous and common social problem faced worldwide. Crimes affect the quality of life ,economic growth and reputation of nation. With the aim of securing the society from crimes, there is a need for advanced systems and new approaches for improving the crime analytics for protecting their communities. We propose a system which can analysis, detect, and predict various crime probability in given region. This paper explains various types of criminal analysis and crime prediction using several data mining techniques.

In many different fields, there is a high demand for storing information to a computer storage disk from the data available in printed or handwritten documents or images so that it can be reutilized later. One simple way is to scan the documents and store them into images but to re-utilize this data for further processing is very difficult e.g. calculating the net amount from scanned bills is a laborious task. It uses two main approaches to accomplish this task: classifying words directly and character segmentation. For the former, it uses Machine Learning technique i.e. Convolutional Neural Network (CNN) with various architectures to train a model that can accurately classify words. For the latter, it uses Long Short Term Memory networks (LSTM) with convolution to construct bounding boxes for each character. It then pass the segmented characters to a CNN for classification, and then reconstruct each word according to the results of classification and segmentation. It uses real time scanning of the evaluation sheet is done and the data is stored in the excel sheet using the proposed feature representation.

Effectively and efficiently learning an optimal kernel is of great importance to the success of kernel method. Along with this line of research, many pioneering kernel learning algorithms have been proposed, developed and combined in many ways. This paper aim to explain the application of Localized Multiple Kernel Support Vector Regression (LMKSVR) to predict the daily stock price of PT. XL Axiata Tbk (EXCL), PT. Indofood Sukses Makmur Tbk (INDF) and PT. Unilever Indonesia Tbk (UNVR) from January 2014 to May 2016. It can be concluded that LKMSVR has good performance to predict daily stock price with Mean Absolute Percentage Error (MAPE) produced all less than 2%.

A car price prediction has been a high-interest research area, as it needed recognizable effort and knowledge of the field expert. This paper mainly focuses on working of three different kind regression algorithms which are used to predict price of a used car. In this project, We have Considered number of distinct attributes which are examined for the reliable and accurate prediction. To build a model for predicting the price of used cars we have used three different kinds of machine learning techniques which comes under supervised machine learning type of algorithm which are linear regression, lasso regression and ridge regression respectively. we have used Python libraries to design GUI for our project and some other machine learning related libraries like Numpy, Pandas, Sklearn etc. we have calculated and compared the accuracies of three machine learning algorithms. The accuracies for linear regression, lasso and Ridge regression were 83.65%, 87.09% and 84.00% respectively. The final main price is predicted according to lasso regression as it gives highest accuracy amongst three different algorithms.

Generalized Information (GI) is a measurement of the degree to which a program can be said to generalize a dataset. It is calculated by creating a program to model the data set, measuring the Active Information in the model, and subtracting out the size of the model. Active Information allows GI to be usable with both exact and inexact models.

In collaborazione con Intech Microwaves-The Round, azienda del Tecnopolo Tiburtino che opera negli ambiti Microonde, Microelettronica e Tecnologie Digitali con la sua divisione “Augmented Communication”, Il Polo Tecnologico ha organizzato il Workshop dedicato a “La realtà aumentata nell’impresa 4.0 - Strategie, opportunità, nuovi modelli di business”.

Il Workshop ha illustrato come le nuove tecnologie digitali possono guidare le aziende all’interno della trasformazione verso l’Impresa 4.0, anche attraverso l’utilizzo di specifiche misure finanziarie agevolative.
Sono stati  presentati anche casi-esempio su Realtà Aumentata, Realtà Virtuale e Realtà Mista e come queste tecnologie possono interagire con l’Intelligenza Artificiale, il  Machine Learning, l’Internet of Things (IoT).

This paper examines the cobalt-doped ceria/reduced graphene oxide (Co-CeO2/rGO) nanocomposite as a supercapacitor and modeling of its cyclic voltammetry behavior using Artificial Neural Network (ANN) and Random Forest Algorithm (RFA). Good agreement was found between experimental results and the predicted values generated by using ANN and RFA. Simulation results confirmed the accuracy of the models, compared to measurements from supercapacitor module power-cycling. A comparison of the best performance between ANN and RFA models shows that the ANN models performed better (value of coefficient of determination >0.95) than the RFA models for all datasets used in this study. The results of the ANN and RFA models could be useful in designing the unique nanocomposites for supercapacitors and other strategies related with energy and the environment.

Decision making both on individual and organizational level is always accompanied by the search of other's opinion on the same. With tremendous establishment of opinion rich resources like, reviews, forum discussions, blogs, micro-blogs, Twitter etc provide a rich anthology of sentiments. This user generated content can serve as a benefaction to market if the semantic orientations are deliberated. Opinion mining and sentiment analysis are the formalization for studying and construing opinions and sentiments. The digital ecosystem has itself paved way for use of huge volume of opinionated data recorded. This paper is an attempt to review and evaluate the various techniques used for opinion and sentiment analysis.

Provision of an accurate and timely disaster relief effort requires an insightful informational coverage of a crisis situation. Social media has a potential of providing a large amount of information generated by people about different aspects of a certain crisis situation. However, emergency-oriented social media utilization is still at an embryonic stage of its understanding. So, in this report 10 research studies were reviewed in an attempt to gain some understanding about this approach. As it appeared, crowdsourced information allows to get a versatile and insightful understanding of a crisis situation, but, as with any Big Data, there are major challenges of making use of high volumes of diverse information, in terms of extraction and representation of relevant and accurate information. Furthermore, other research studies proposed extraction/representation techniques which were primarily aimed at thematic and geographical clustering of information. As it appeared from the real cases of social media application during emergencies, there is still no standardized approach on how such information can be handled, but in those cases, emergency responders still managed to successfully apply their own methods of primarily geographical and thematic clustering of information, while still being greatly challenged by overwhelming amount of user-generated information. Thus, the success could perhaps be explained by social media being used primarily as a complement/enhancement of other already proved to be effective techniques. Further exploration would be needed in order to gain a more complete understanding of the current state emergency-oriented social media utilization, which could later be used as a theoretical framework for a new research study in this area.

Sentiment analysis acquired a great area of attention in the microblogging websites and analysis of sentiment is a practice of categorization and identification of opinions that are articulated as speech, text, database sources and tweets to detect if opinion is negative, positive or neutral. The challenge lies in determining sentiment from the tweets due to the unique characteristics of Twitter data. This paper presents an approach for sentiment analysis by adapting a Hadoop framework and deep learning classifier. The Hadoop cluster is used for the distribution of data for extracting the features. Then, the significant features are extracted using the twitter data. The deep learning classifier, namely deep recurrent neural network classifier is used assign a real-valued review to each input twitter data thus, classifying the input data into two classes, such as positive review and negative review. The analysis of the performance is done using metrics like, classification accuracy, sensitivity and specificity. In contrast to classical strategies, the proposed method offered improved classification accuracy of 0.9302, better sensitivity of 0.9404 and high specificity of 0.9157, respectively.

Analysis and study of sleep patterns is very important for evaluating the sleep quality and sleep disorders. The sleep stage scoring has been done traditionally by human expert technicians over many years and needs a significant amount of time and human effort. The aim of this project is to build a system to generate the sleep scores from the polysomnographic (PSG) sleep data automatically. The proposed system first leverages the Big Data techniques to extract and transform the vast amount of PSG sleep data in a distributed environment. The transformed data, extracted from the single most important EEG channel is then used to train the machine learning models like Support Vector Machine, Random Forest Classifier and deep learning models like MLP and GRU, in order to predict the sleep stages. The study also compares the performance of these models and discusses the results of different approaches. The current best performance model is Variable Recurrent Neural Network with test accuracy score of 77.7%.

Content based image retrieval (CBIR), also known as query by image content (QBIC) is the application of computer vision to the image retrieval problem, that is, the problem of searching for digital images in large databases. Users of image databases often prefer to
retrieve relevant images by categories. Unfortunately, images are usually indexed by content or low level features like color, texture and shape, which often fail to capture high level concepts well. Again the high level concepts or semantics can vary from user to user. To address these issues, relevance feedback has been extensively used to associate low level image features with high level concepts. Because it is hard to define what will be the most similar images of a query image as it varies from a user to another, taking user judgement on the retrieved images and then refining the search result based on user’s region of interest has become very successful. Many researchers have addressed the issues of Content based Image retrieval with relevance feedback and contributed in this field to a considerable extent. Still there are scope and need of much improvement. Among all existing relevance feedback approaches, query point movement and feature re-weighting have been proven to be suitable for large-scaled image databases with high dimensional image features. In this thesis, we present a query point movement approach using both relevant and irrelevant images to catch the user semantics. The relevance feedback strategy is simple and straightforward. The basic approach adopted here is to move the query point to the direction of user intention. The user only selects the images that he/she thinks relevant and
ignores the irrelevant ones. The feedback provided by the user are categorized into positive and negatives images both of which categories are used in determining the query point for further iterations. During experimentation it is observed that our system can efficiently move towards the user context with minimum possible number of iterations comparing to trivial methods.