Data Scientist
Data Science Trainee at AlmaBetter with an interest in analyzing raw data, statistics and hands-on experience in building machine learning models.
Overview
2
2
years of professional experience
6
6
years of post-secondary education
10
10
Certifications
Work History
Production Officer
Alok Industries Limited
10.2020 - 12.2020
Analyzed the root cause of the defects occurring in yarn.
Planned the daily production effectively and scheduled the shift to achieve the daily
production targets.
Production Engineer
Kristeel Shinwa Industries Limited
07.2018 - 10.2020
Analyzed the SOPs to formulate necessary ISO documentations and plannned the
Monthly and Daily Production.
Lead a team of 15 members to achieve the production targets.
Worked with the R &
D department to effectively reduce the cost of production.
Book Recommender System
AlmaBetter Verified Project
09/2021,
Tags: EDA, Feature Engineering, KNN, Collaborative Filtering, Content Based Filtering, NLP, TFIDF,
SVD, Recall
Developed different recommender systems for recommendations of books to users,
using explicit and implicit ratings of books and popularity-based, collaborative, and
content-based filtering.
Extracted new information of the books using Google API and vectorized it using
TFIDF.
Used models like KNN and Singular Value Decomposition to make
recommendations based on user-user and item-item similarities.
Obtained a recall of 90% using collaborative filtering.
Bankruptcy Prediction
AlmaBetter Verified Project
09/2021,
Tags: EDA, Feature Selection, Quasi, L1 Regularization, Random Forest , PCA, Isolation Forest,
Resampling, Easy Ensemble, Gaussian Naive Bayes, XGBoost
Developed a model which captures the bankruptcy patterns among the companies
in industries.
Reduced the features using different techniques like Random Forest Feature
Importance, p_values, VIF, Quasi Feature Selection and L1 Regularization.
Built multiple models with the combinations of these features considering different
use cases.
Built a Random Forest Model which gave a recall of 94%, a KNN model which gave
a precision of 74%, and a Gaussian Naive Bayes Model which gave an F1 score of
51%.
Certification
Discrete RV, Continuous RV, Probability Distributions