Software Engineer @ apna | IIT KGP'22

Saurav .  • Peerlist

•Formulated an expression for price optimization in the presence of cross channel interaction in demand and supply
•Applied pricing policy based on the partition of store inventories that estimated the way shared resources are utilized
•Solved above computationally tractable mixed-integer linear programming problem through IBM CPLEX Optimization Studio
•Applied Metaheuristic algorithms which reduced the computational time thereby improving the scalability of the problem

Dynamic Pricing in Omnichannel Inventories

•Developed a decision-making tool for dynamic bank lending decisions to construct optimal loan portfolios in a credit crunch environment
•Applied Genetic Algorithm based framework (GAMCC) to maximize the net profit of the bank and minimize loss due to credit default
•GAMCC reduces the average computational time from 1.03 second (in traditional linear programming) to 0.54 second, for 400 customers

Optimization of bank lending decisions

•Built a multi class classification model based on the Non-Linear SVM algorithm for classifying glass type by adopting the OVA strategy
•Verified various classifier models using several kernel functions (Gaussian, RBF, Polynomial) and by tuning their respective parameters
•Adopted k-fold cross-validation technique by stratified sampling for an unbiased evaluation of classifier models on a limited data sample
•Measured performance in terms of Predictive accuracy, F1score, Precision and Recall and constructed the respective ROC curve