The introductory week will look into forward feed neural networks. The feedforward neural network was the first and simplest type of artificial neural network devised. In this network, the information moves in only one direction forward-from the input nodes, through the hidden nodes (if any) and to the output nodes. There are no cycles or loops in the network. We will also look into -

• Neural Network
• Different architecture of Neural Network
• Importance of Neural Network
• Hyperparameters in Neural Network
• Different types of Gradient descent methods

In the second week, we will look into activation functions. Activation functions are mathematical equations that are used to determine the output of a neural network. The function is attached to each neuron in the network, and is used to determine whether it should be activated or not, based on whether each neuron's input is relevant for the model's prediction. In activation functions, we will look into- 

• Conic sections
• Hyperbolic trigonometric functions
• Sigmoid activation function
• Tanhx activation function
• Relu activation function
• Softmax activation function

Deep learning is a branch of machine learning which is based on artificial neural networks. A neural network is a mimic of the human brain, so in short deep learning is also a mimic of human learning techniques. In deep learning, the software engineer doesnt program everything, instead we focus on developing our neural network. In this week, we will be looking into-

• Terminologies in deep learning
• Nomenclature
• Order of vectorized forms
• Forward propagation derivation with 1 layer
• Back propagation derivation with 1 layer
• Batch size, Iteration and Epoch

The performance of a machine learning model can be characterized in terms of the bias and the variance of the model. A model with high bias makes strong assumptions about the form of the unknown underlying function that maps inputs to outputs in the dataset, such as linear regression. A model with high variance is highly dependent upon the specifics of the training dataset, such as unpruned decision trees. So, what are bias and variance?- Bias is the simplifying assumptions made by the model to make the target function easier to approximate and variance is the amount that the estimate of the target function will change given different training data. So, in this week we will look into different topics related to evaluation of models-

• Underfitting
• Overfitting
• Lasso regularization
• Ridge regularization
• Elastic Net regularization

Ensemble methods are meta-algorithms that combine several machine learning techniques into one predictive model in order to decrease variance, bias, or improve predictions.

• Sparse and convex functions
• Bagging to avoid overfitting
• Boosting to avoid underfitting
• Stacking to avoid underfitting

Regularization is a set of techniques that can prevent overfitting in neural networks and thus improve the accuracy of a Deep Learning model when facing completely new data from the problem domain. And Optimizers are algorithms or methods used to change the attributes of your neural network such as weights and learning rate in order to reduce the losses. In this week we will look into the different types of regularisers and optimisers- 

• Frobenius norm regularization
• Data augmentation
• Early stopping
• Adam optimizer
• Tensorflow 2.0

In deep learning, a convolutional neural network is a class of deep neural networks, most commonly applied to analyzing visual imagery. In this week we will look into more topics related to CNN such as-

• Basics of CNN
• Edge detection
• Padding
• Stride
• Simple CNN
• Difference between CNN & ANN

This week is a continuation of the previous week’s topic. Here we will look into more topics related to CNN.

• Pooling layers
• Transfer learning
• Examples on CNN architecture
• Combination of different Neural network architecture
• CNN in python

A recurrent neural network (RNN) is a class of artificial neural networks where connections between nodes form a directed graph along a temporal sequence. This allows it to exhibit temporal dynamic behavior. Derived from feedforward neural networks, RNNs can use their internal state (memory) to process variable length sequences of inputs. This makes them applicable to tasks such as unsegmented, connected handwriting recognition or speech recognition.

• RNN Model
• Different types of RNN
• Gradients in RNN
• Back propagation
• Difference between RNN & ANN

This week is a continuation of the previous week’s topic. Here we will look into more topics related to RNN.

• Gated Recurrent Unit (RNN)
• Long short term memory (LSTM)
• Bidirectional RNN
• RNN Implementation in Python

Natural Language Processing(NLP) refers to computer systems designed to understand human language. Human language, like English or Hindi consists of words and sentences, and NLP attempts to extract information from these sentences. Word embedding is any of a set of language modeling and feature learning techniques in natural language processing where words or phrases from the vocabulary are mapped to vectors of real numbers.

• Stop words
• Stemming
• Lemmatization
• Word2vec
• Implementation of word2vec in python

In this week, we will be performing an end-to-end ML/DL project for machine failure.

• Descriptive Analytics
• Diagnostic Analytics
• Predictive Analytics
• Prescriptive Analytics