My name is
Abhishek Chaudhary
A data scientist enthusiast who believes that the power of data will revolutionize the world
My name is
Currently
Skill-Set
Project-1
Covid Prediction
A projct to detect Covid/Viral Pneumonia cases using Chest X-ray with the help of tensorflow and keras, I have used Convolution Neural Network having VGG-16 architecture algorithm, I also have used callback function to get a decent acuracy. And after building the model, i have dumped into a .h5 file to use it again for prediction
Project-2
Divorce Prediction
A project to predict divorce using a set of questionnaire, the dataset was having a set of 54 questions which was reduced to 17 with the help of Recursive feature elimination with cross validation. With using that i have found out that 17 no. of questions is the sweet-spot. After that i got a accuracy of 97.60% of test set.
Project-3
Predictive maintenance
My M.Tech final year project, the aim of the project is to predict the failure and use this insight to schedule maintenance before the actual failure occurs. Here there are two target variables, one of it was of binary (Failure) class and one of it was of multi class (Type of failure) feature variable. I have tried different types of ML algorithm, and found that RandomForest Classifier was giving the best performance with almost 99.5% of accuracy. There is cache in this project, that as there are two target variables, so i have used two RandomForest Classifier to predict both the target variable, and i also taken care that none of the variable is used to predict the other target variable, which will leads to memory leak problem.
Project-4
Sentiment analysis of my college
The aim of the project is to perform sentiment analysis of NIT Trichy, at first I have performed web-scrapping to get the reviews from different educational website (eg., Career360) using BeautifulSoup library by html parser, HTML tags has also been carefully taken care of and after that exported into a xlsx file and saved. and after that i performed various NLP techniques like Normalizing and Cleaning, Tokenization, Stemming, Removal of stop words and then performed sentiment analysis and generated sentiment polarity and subjectivity in the range of [0,1] and after that i have also created a dashboard using PowerBi.
Peoject-5
Credit Card Fraud detection
Aim of this was to detect credit card fraduelent cases, in this project i have used logistic regression cause of easy tuning of its threshold which is required to reduce the false negative cases. I have also performed feature engineering and used ROC curve to decide the best threshold for logistic regression algorithm.
Project-6
Ineternship at Henry Harvin
During this internship i was assigned to perform sentiment analysis on reviews of Henry Harvin Educational institute from various websites, eg., Youtube, Twitter, Trustpilot, Career360. Performed web-scrapping with the help of BeautifulSoup library and after that i performed various NLP techniques like Normalizing and Cleaning, Tokenization, Stemming, Removal of stop words and then performed sentiment analysis and generated sentiment polarity and subjectivity in the range of [0,1] and after that i have also created multiple dashboard for different courses and mentor of the institute using PowerBi.