GRU are a kind of Neural Networks which are designed to overcome some issues with RNN. There are two primary issues with recurrent neural networks Calculations of gradients either fail or explode. Gradient calculations are expensive Gradient clipping is a solution to the expanding gradient problem, and other topologies like the gated recurrent unit (GRU) […]
machine learning
Stock Price Prediction On Commercial Data Using LSTM
LSTMS are a variety of RNNs but lets start with RNNs. RNNs are unable to remember long-term dependencies in time series data because of the vanishing gradient issue. An RNN version called LSTM was created to deal with this problem. Similar to RNN, LSTM features a hidden state that functions as short-term memory. Additionally, it […]
DBSCAN Clustering Algorithm
DBSCAN is an example of clustering algorithm. Clustering is the process of organizing a collection of concrete or abstract things into classes of related objects. A cluster is a group of data objects that are distinct from the objects in other clusters yet comparable to one another within the same cluster. There are various example […]
Vector Space Model in NLP
Lets begin with fundamentals for vector space model. All NLP applications require a basic understanding of how word vectors are represented in natural language processing. Because the first step in creating any NLP model is to express the word as some kind of number or coding. For instance, a word can be represented as a […]
Basics of Natural Language Processing (NLP) Pipeline
Images in this NLP blog are borrowed from lecture note of Prof. Bal Krishna Bal. The end-to-end structure that orchestrates the flow of data into and output from a NLP model is known as a natural language processing pipeline (or set of multiple models). It covers the input of the raw data, the features, the […]
Approaches for Natural Language Processing (NLP)
Several widely used NLP algorithms Images in this blog are borrowed from lecture note of Prof. Bal Krishna Bal. Not only are there many algorithms for machine learning tasks, but also for natural language processing job. The choice of which algorithm is best for a given problem cannot be made with absolute certainty. After determining […]
Multilayer Perceptron: Solving XOR Problem
Multilayer Perceptron: Solving XOR Problem from Scratch in Python In this blog we are going to explore how non-linear problem like XOR can be solved using multi layer perceptron. We have already tried how to apply multi layer perceptron on majority function please have a look here. We all are familiar with single layer perceptron […]
Multilayer Perceptron Using Majority Function From Scratch
Multilayer Perceptron (MLP) We all know that single layer perceptron are commonly used to classify problems that are linearly separable. If we choose a single layer perceptron for a non-linearly separable problem, the results may not be successful. As a result, we must look for an alternative solution to a non-linear problem, and one such […]
Beyond and Within EDA: Taking EDA into Modelling
Beyond and Within EDA Introduction This blog is the continuation of the previous blog post A General Way of Doing EDA. Please follow that before reading this blog. Once we got the knowledge of the data like its properties and features, we can move ahead by taking that knowledge to make some sort of inference. […]
Single Layer Perceptron From Scratch
This blog contains the introduction to perceptron using Python. Artificial Neuron Network The term "neural network" refers to an artificial neural network (ANN) (NN), which is a computational paradigm inspired by the way the human brain or nervous system performs computation. Perception, pattern recognition, motor control, and other computations are all performed by the brain, […]