Yash Saini's Projects
This mini project uses a pre given dataset of a bank (10,000 rows) having various independent and the dependent variable of binary type( whether the customer leaves the bank or not). ANN has been used and an accuracy of 84.2% has been achieved.
As per the Deep learning course A-Z , a movie recommender system was created for the MovieLens dataset.
Using Restricted Boltzmann machines , a movie recommender system was created
This application caters to building a full stack application depicting the features of snippets of notes saved to the cloud.
This project contains the backend work of the cloud_notebook
Convolutional Neural Network program helps in classifying the data consisting of 10,000 images of Cats and dogs.
This project illustrates the use of unsupervised ML technique called Self Organizing Maps to detect frauds committed in getting a credit card issued. The dataset has been downloaded from UCI machine learning repository.
The project caters to finding the 5 sentiment scores (Happiness,Fear,Anger, Disgust,Sadness) of each tweet using a described formula.
This research deals with analyzing the hate speech with a novel method of combining LDA with SOM. Experimentally no of topics derived from the dataset after the application of LDA was 10. Dominant topic was calculated for each data value. This data was fed to a SOM network initially assigned with random weights. The learning rate for SOM was 0.5 and number of iterations were 100. The clusters were formed as a result and the map was displayed.
This repository contains the jupyter notebook catering to demonstrating the image augmentation techniques using pytorch.
Junction is a software to manage proposals, reviews, schedule, feedback during conference.
Simulation-based Lidar Super-resolution for Ground Vehicles
The following code depicts text generation using LSTM. The dataset used is Shakespeare's Macbeth. 50 characters are taken and 51st character is predicted using this model. 1 epoch cycle with 0.1 million entries fed to train the network.
The project is based on using supervised machine learning techniques (KNN,SVM and xgboost) on wisconsin dataset of breast cancer, the accuracy and other metrics of the techniques are pit against each other.
This project makes use of various ensemble learning techniques for example bagging, boosting ,gradient boosting machine and extreme gradient boosting machine on an HR dataset attached herewith. The measures of performance for each technique is calculated.
All Algorithms implemented in Python
The project analyzes the stocks of three major car companies , compares them , plots graphs and calculates returns & cumulative returns on them.
The repository contains some of the Python Programs
This repository caters to discussing about various tips and tricks which could be applied on any react project. Discussions revolve around essential react concepts and their practical applications
The project 'Sentiment Analysis of top colleges in India using Twitter Data', extracts tweets from the twitter , preprocesses them , assigns sentiment scores and then using different ML techniques , their accuracies are compared.
RNN is used here to predict the trends using historical stock data obtained through yahoo finance