Scribd Logo
Upload

Welcome to Scribd!

  • 116 views

    CCS355 SET1 Anna University Lab Manual Question Set

    Uploaded by

    Pandi Selvi

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd

    CCS355 SET1 Anna University Lab Manual Question Set

    Uploaded by

    Pandi Selvi
    116 views3 pages

    Original Title

    CCS355 SET1 Anna university lab manual question set

    Copyright

    © © All Rights Reserved

    Available Formats

    PDF, TXT or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd
    Download as pdf or txt
    116 views3 pages

    CCS355 SET1 Anna University Lab Manual Question Set

    Uploaded by

    Pandi Selvi

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd
    Download as pdf or txt
    You are on page 1of 3

    B.E / B.Tech.

    PRACTICAL END SEMESTER EXAMINATIONS, NOVEMBER/DECEMBER 2023

    Fifth Semester

    CCS355 - NEURAL NETWORKS AND DEEP LEARNING

    (Regulations 2021)

    Time: 3 Hours Answer any one Question Max. Marks 100

    Aim and Procedure Program Calculation Viva-Voce Record Total


    & Results

    20 30 30 10 10 100

    1. Implement mathematical operations addition, subtraction, multiplication and dot product, length,
    concatenation of vectors using Tensorflow

    2. Use AutoMPG data set from the UCI Machine Learning Repository and build the models to
    predict the fuel efficiency of the automobiles (Use Keras API)

    3. Load the MNIST dataset from Tensorflow and implement a multilayer perceptron in
    Tensorflow environment. Initialize the input vector, Train the network weights, and Make
    predictions with the perceptron.

    4. Use CIFAR-10 dataset and implement Feed-Forward Network in TensorFlow/Keras

    5. Classify CIFAR images using Convolutional Neural Networks in TensorFlow/Keras

    6. Using the VGG-16 Deep Convolutional Network as a Transfer Learning framework, Classify the
    Fruits dataset into 3 classes namely Apple, Banana and Orange.

    7. Create a feedforward neural network to classify images of handwritten digits from the MNIST
    dataset into one of the ten classes (0-9). Use softmax activation and categorical cross-entropy
    loss.

    8. Implement the image classification using the VGG-16 Deep Convolutional Network used as a
    Transfer Learning framework in the Fruits dataset to classify the fruits into Apple, Banana and
    Orange

    Page 1 of 3
    9. Implement a transfer learning solution for image classification problems using Pretrained model
    on Keras

    10. Import IMDB dataset from keras and perform Sentiment Analysis with an Recurrent Neural
    Networks (RNN).

    11. Implement a simple Long Short Term Memory autoencoder with the help of Keras and python
    (Can use for example S&P500 Daily Prices dataset with Date and Closing Price). Reduce
    overfitting if necessary.

    12. Generate images and discriminate the true and fake images for MNIST data set using
    Generative Adversarial Networks (GANs)

    13. Build a CNN for image classification using the CIFAR-10 dataset, which consists of 32x32 color
    images across ten classes. Design a CNN architecture with convolutional layers, pooling layers,
    and fully connected layers for image classification.

    14. Implement a RNN to generate text character by character. Train the network on a dataset of text
    and generate sequences of text. Train an RNN on a corpus of Shakespearean text to generate
    new text passages in the style of Shakespeare

    15. Use LSTM networks for sentiment analysis on a dataset of movie reviews, predicting positive,
    negative, or neutral sentiment.

    16. Given a dataset of two features (X1 and X2) and binary labels (0 or 1), create a neural network
    with a single neuron to perform binary classification. Train the network using the sigmoid
    activation function and binary cross-entropy loss.

    17. Create a feedforward neural network to classify images of handwritten digits from the MNIST
    dataset into one of the ten classes (0-9). Use softmax activation and categorical cross-entropy
    loss.

    18. Load a pre-trained BERT model and fine-tune it for sentiment analysis on a dataset of movie
    reviews, predicting positive or negative sentiment

    19. Create a convolutional autoencoder to compress and decompress images. Train the
    autoencoder on a dataset of grayscale images.

    Page 2 of 3
    20. Optimize the hyperparameters of a neural network for image classification to achieve the
    highest accuracy. Experiment with different learning rates, batch sizes, and architectures to
    improve the accuracy of an image classifier

    Page 3 of 3

    You might also like