RTMNU Final CSE (DS) syllabus

RTMNU · Final Year B.Tech Computer Science & Engineering (Data Science) · CBCS

Deep Learning

Elective IV · Semester VII · 3 credits · College 30 + university 70

One of the Elective IV options. A different syllabus from the CSE (AI & ML) Deep Learning; Unit 4 prints no hours.

  1. Unit 1: Deep Architectures

    7 hours

    Why deep architectures and where they apply; hyperparameters — encoding, layers, loss functions, learning rate, momentum and optimisation, regularisation and dropout, batch norms; the vanishing-gradient problem and how to mitigate it.

    Examined in the 70-mark end-semester paper

    • Hyperparameters in deep learning
    • Loss functions
    • Regularisation and dropout
    • Vanishing and exploding gradients
  2. Unit 2: Deep Learning in Computer Vision

    7 hours

    The origins of CNNs, convolutional neural networks, fine-tuning CNNs and popular CNN architectures.

    Examined in the 70-mark end-semester paper

    • Convolution and feature maps
    • CNN architectures: LeNet to ResNet
    • Transfer learning
  3. Unit 3: Deep Learning for NLP

    7 hours

    Traditional and deep-learning NLP and vector representations; deep learning for text — recurrent networks and LSTMs — and applications.

    Examined in the 70-mark end-semester paper

    • Word embeddings: Word2Vec and GloVe
    • Recurrent neural networks for NLP
    • LSTM
  4. Unit 4: Multimodality and Deep Reinforcement Learning

    Multimodal learning and its challenges, image captioning, visual question answering and multi-source self-driving; deep reinforcement learning and implementing it.

    Examined in the 70-mark end-semester paper

    • Multimodal models
    • Image captioning and video to text
    • Visual question answering
    • Deep reinforcement learning
  5. Unit 5: Deep Learning Hacks and Trends

    7 hours

    Preparing data, training tricks, fine-tuning and model compression; recent models and novel applications.

    Examined in the 70-mark end-semester paper

    • Data augmentation
    • Fine-tuning pretrained and masked language models
    • Model compression and pruning

Unit contents are summarised from the official RTMNU syllabus . AiCanCode.org is not affiliated with Rashtrasant Tukadoji Maharaj Nagpur University (RTMNU).