0% found this document useful (0 votes)
4 views7 pages

Deep Learning Lab

The document outlines various deep learning projects, including building neural networks for tasks such as solving the XOR problem, recognizing handwritten digits and characters, face recognition, language modeling, sentiment analysis, parts of speech tagging, machine translation, image augmentation, and predicting disease outcomes and stock prices. Each project includes a grading rubric for aim, program, results, viva-voce, record, and procedure, totaling 100 points. The projects utilize different datasets and neural network architectures like DNN, CNN, RNN, LSTM, and GAN.
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
4 views7 pages

Deep Learning Lab

The document outlines various deep learning projects, including building neural networks for tasks such as solving the XOR problem, recognizing handwritten digits and characters, face recognition, language modeling, sentiment analysis, parts of speech tagging, machine translation, image augmentation, and predicting disease outcomes and stock prices. Each project includes a grading rubric for aim, program, results, viva-voce, record, and procedure, totaling 100 points. The projects utilize different datasets and neural network architectures like DNN, CNN, RNN, LSTM, and GAN.
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd

Build a Deep Neural Network (DNN) to solve the XOR problem.

Use a dataset containing


input-output pairs (0, 0), (0, 1), (1, 0), and (1, 1). Use different activation functions in each
layer and demonstrate how the network learns XOR logic.
Aim/ Program (30) Results (30) Viva-Voce(10) Record (10) Total
Procedure(20) (100)

INTERNAL EXAMINER EXTERNAL EXAMINER

Build a Convolutional Neural Network (CNN) to recognize handwritten digits. Utilize the
MNIST dataset, a collection of 28x28 grayscale images of handwritten digits (0-9), for training
and testing your CNN.
Aim/ Program (30) Results (30) Viva-Voce(10) Record (10) Total
Procedure(20) (100)

INTERNAL EXAMINER EXTERNAL EXAMINER

Implement a Convolutional Neural Network (CNN) for recognizing handwritten characters. Utilize the
EMNIST dataset, which contains handwritten letters and digits.

Aim/ Program (30) Results (30) Viva-Voce(10) Record (10) Total


Procedure(20) (100)

INTERNAL EXAMINER EXTERNAL EXAMINER


Develop a CNN-based face recognition system. Use a simplified dataset containing grayscale
face images of a few individuals for training and testing your model.
Aim/ Program (30) Results (30) Viva-Voce(10) Record (10) Total
Procedure(20) (100)

INTERNAL EXAMINER EXTERNAL EXAMINER

Develop a CNN-based face recognition system. Use a simplified dataset with grayscale face
images of multiple individuals for training and testing your model.
Aim/ Program (30) Results (30) Viva-Voce(10) Record (10) Total
Procedure(20) (100)

INTERNAL EXAMINER EXTERNAL EXAMINER

Create a character-level language model using Recurrent Neural Network (RNN). Use a small
corpus of text (such as Shakespearean plays or poems) for training and generating new text.
Aim/ Program (30) Results (30) Viva-Voce(10) Record (10) Total
Procedure(20) (100)

INTERNAL EXAMINER EXTERNAL EXAMINER


Create a word-level language model using Recurrent Neural Network (RNN). Use a dataset of
sentences or paragraphs from literature.
Aim/ Program (30) Results (30) Viva-Voce(10) Record (10) Total
Procedure(20) (100)

INTERNAL EXAMINER EXTERNAL EXAMINER

Perform sentiment analysis on movie reviews using Long Short-Term Memory (LSTM)
networks. Use the IMDB dataset, consisting of movie reviews labelled as positive or negative,
for training and testing your LSTM model.
Aim/ Program (30) Results (30) Viva-Voce(10) Record (10) Total
Procedure(20) (100)

INTERNAL EXAMINER EXTERNAL EXAMINER

Perform sentiment analysis on customer reviews using Long Short-Term Memory (LSTM)
networks. Use the amazon dataset, consisting of product reviews labelled as positive or
negative, for training and testing your LSTM model
Aim/ Program (30) Results (30) Viva-Voce(10) Record (10) Total
Procedure(20) (100)

INTERNAL EXAMINER EXTERNAL EXAMINER


Perform sentiment analysis on customer reviews using Long Short-Term Memory (LSTM)
networks. Use the Flipkart dataset, consisting of product reviews labelled as positive or
negative, for training and testing your LSTM model.
Aim/ Program (30) Results (30) Viva-Voce(10) Record (10) Total
Procedure(20) (100)

INTERNAL EXAMINER EXTERNAL EXAMINER

Implement parts of speech tagging using a Sequence-to-Sequence architecture. Use a dataset


containing sentences with corresponding parts of speech tags for training and evaluating your
model.
Aim/ Program (30) Results (30) Viva-Voce(10) Record (10) Total
Procedure(20) (100)

INTERNAL EXAMINER EXTERNAL EXAMINER

Implement parts of speech tagging using a Sequence-to-Sequence architecture. Use a dataset


containing sentences with corresponding parts of speech tags. Use encoder-decoder structure
and attention mechanisms in sequence-to-sequence models.
Aim/ Program (30) Results (30) Viva-Voce(10) Record (10) Total
Procedure(20) (100)

INTERNAL EXAMINER EXTERNAL EXAMINER


Build an English to Tamil machine translation system using an Encoder-Decoder model. Use
a small parallel corpus of English and Tamil sentences for training and testing your translation
model.
Aim/ Program (30) Results (30) Viva-Voce(10) Record (10) Total
Procedure(20) (100)

INTERNAL EXAMINER EXTERNAL EXAMINER

Build a machine translation system to translate English sentences to French using an Encoder-
Decoder model. Use a parallel corpus containing English and French sentence pairs.
Aim/ Program (30) Results (30) Viva-Voce(10) Record (10) Total
Procedure(20) (100)

INTERNAL EXAMINER EXTERNAL EXAMINER

Implement image augmentation using Generative Adversarial Networks (GANs). Use a dataset
of natural images (such as animals or objects) to generate augmented versions of the images for
training other models.
Aim/ Program (30) Results (30) Viva-Voce(10) Record (10) Total
Procedure(20) (100)

INTERNAL EXAMINER EXTERNAL EXAMINER


Implement image augmentation using Generative Adversarial Networks (GANs). Use a
dataset of natural scenes or objects. Show how GANs can generate realistic variations of
existing images and improve the diversity of the dataset for training image recognition
models.
Aim/ Program (30) Results (30) Viva-Voce(10) Record (10) Total
Procedure(20) (100)

INTERNAL EXAMINER EXTERNAL EXAMINER

Develop a deep learning model to predict disease outcomes based on patient data.
Aim/ Program (30) Results (30) Viva-Voce(10) Record (10) Total
Procedure(20) (100)

INTERNAL EXAMINER EXTERNAL EXAMINER

Create a deep learning model to predict stock prices based on historical data.
Aim/ Program (30) Results (30) Viva-Voce(10) Record (10) Total
Procedure(20) (100)

INTERNAL EXAMINER EXTERNAL EXAMINER


Analyze customer behavior using deep learning to improve product recommendations.
Aim/ Program (30) Results (30) Viva-Voce(10) Record (10) Total
Procedure(20) (100)

INTERNAL EXAMINER EXTERNAL EXAMINER

Implement a Deep Neural Network (DNN) to solve the XOR problem. Use a small dataset
containing input-output pairs (0, 0), (0, 1), (1, 0), and (1, 1) for training and testing your
model.
Aim/ Program (30) Results (30) Viva-Voce(10) Record (10) Total
Procedure(20) (100)

INTERNAL EXAMINER EXTERNAL EXAMINER

You might also like