Deep Learning Training In Hyderabad

Datajango Provides end to end Deep Learning Course with 100% Industry Ready code examples in Hyderabad. The course we will cover all necessary concepts to make a successful Data Scientist. The concepts we cover are Deep Neural Networks, Convolutional Neural Networks, Computer Vision, Natural Language Processing& Information Retrieval with Gensim, SpaCy and NLTK, Recurrent Neural Networks for Text Analytics.

Deep Learning, Tensorflow, and DNN Course Syllabus

Section – I (Deep Neural Networks) – 20 hours

• • Introduction to Neural Networks
  • Linear Regression Gradient Descent (Batch, Stochastic and Mini-Batch)
  • Logistic/Sigmoid neuron
    • Forward propagation
    • Back Propagation
  • Neural Network Architecture
    • Layers of a Deep Neural Network
    • Back Propagation
    • Activation Functions (Sigmoid, Tanh, ReLU, Leaky ReLU, Softmax)
  • Introduction to TensorFlow
    • Construction Phase
      • tf.Variable
      • tf.constant
      • tf.placeholder
      • Tensor reshape, slice, typecast
      • Variable collections – Global, Local, Trainable
      • Initializing Variables
    • Execution Phase
    • Linear Regression with TensorFlow

Use Case: Build a handwritten digit recognition model with TensorFlow

• Regularizing Deep Neural Networks
  • l1, l2 regularization
  • Dropout regularization
• Vanishing & Exploding Gradients
  • Weight initializations (He/Xavier initialization)
• Algorithm Optimizers
  • Momentum – Exponentially weighted moving average
  • Gradient Descent with Momentum
  • Gradient Descent with RMSProp (Root Mean Squared Propagation)
  • Gradient Descent with ADAM (Adaptive Momentum Estimation)
  • Batch Normalization

Convolutional Neural Networks – Computer Vision

Section – II (Convolutional Neural Networks) – 15 hours

• Introduction to CNN (Convolutional Neural Networks), Computer Vision
• Convolution and Edge detection
• Padding, Striding Convolutions
• Convolution Neural Network
• • Edge Detection
  • Padding
  • Stride
  • Pooling
• ResNets (CNN build with Residual Block)
• Inception Network (filter size, pooling, stride all combined layer)
• Data Augmentation
• Transfer Learning
• Computer Vision
• • Object Location
  • Intersection over Union
  • Anchor Boxes
  • Normax Suppression
  • YOLO Algorithm
  • Object Detection
  • Face Verification

Natural Language Processing, IR and Text Analytics

Section – III (Natural Language Processing & Information Retrieval with Gensim, SpaCy, and NLTK)– 20 hours

• Text Processing with Python (Regular Expressions)
• Introduction to NLP – NLTK, Stanford Core NLP
  • Text Normalization
    • Tokenization
    • Case folding
    • Synonyms, Homonyms
    • Spelling mistakes
    • Stop words
    • Stemming
    • Lemmatization
  • What is Text Corpus?
  • Understanding different corpora in NLTK
• Basic Sentiment Analysis Model (IMDB dataset)
  • Basic feature extraction using CountVectorizer
  • Build Sentiment Classifier
• Information Retrieval (IR)
  • Term-Document incidence matrix
  • Inverted Index
  • Handling Phrase Queries (IR)
    • Biword index
    • Positional index
  • Spelling Correction
    • SoundX algorithm
    • Isolated words
      • Edit Distance
      • Weighted edit distance
      • N-Gram overlap (Jaccard coefficient)
    • Context-sensitive search
  • Document search and Rank Retrieval model
    • Term Frequency, Weighted Term Frequency, Inverse Document Frequency
    • TF-IDF Scoring
    • Euclidian distance
    • Cosine similarity
• Sentiment Classification with TF-IDF (IMDB dataset)
• Sentiment Classification with Hashing Vectorizer (IMDB dataset)
• Word Embeddings

Recurring Neural Networks

Section – IV (Recurrent Neural Networks for Text Analytics) – 15 hours

• Recurrent Neural Networks
• Bidirectional Recurrent Neural Networks
• Gated Recurrent Units (GRU)
• Long short-term memory (LSTM)
• Autoencoders
• Time series (Stock price prediction), Language Generation (Sequence to Sequence model)