Pankaj Kumar

PANKAJ KUMAR (+91)- |-|- EDUCATIONAL QUALIFICATIONS Year- Degree BS - MS (Dual): Mathematics & Scientific Computing Minor : ML, Computer System; Dept. of C.S.E. Class XII (AISSCE) Class X (AISSE) Institute CGPA/% Indian Institute of Technology, Kanpur BVN, Jehanabad BVN, Jehanabad PG: 9.2/10.0 UG: 6.1/10.0 81.2 % 10/10.0 DATA SCIENCE PROJECTS Taxi Fare Prediction | Course Project, Prof. Faiz Ahmed, IIT Kanpur May’19-July’19 Objective  To design a model which can predict the taxi fare based on various factors, with an optimal accuracy  Performed exploratory data analysis on the raw data using various data visualization tools and techniques Approach  Employed Linear regression, Lasso, Random Forest and Artificial Neural Network models on the pre-processed data  Engineered new features to improve RMSE and Variance values of the baseline models used  Model predicted fare with an average error of 5% when compared to a taxi fare predictor device Results Fake News Detection | Prof. Arnab Bhattacharya, IIT Kanpur Jan’19 –May’19  Designed a model to classify a given article as real or fake and to act as a filter for future news articles  Performed pre-processing, tokenization, word stemming to convert text documents into features vector using NLTK  Used feature set as a body text, title, body + title, and applied Logit, Naïve Bayes, ANN, and Support vector machine for classification  Compared results obtained from different classification technique, out of which SVM gave an average accuracy of 93% Text Emojification | Online Project, Prof. Andrew Ng, Coursera Apr’19 –June’19  Build a model to represent a given text or line by an equivalent emoji representation  Employed a multi-layer unidirectional Long short term memory sequence model after word2vec embedding  Implemented above model using Keras, used categorical cross entropy loss with an Adam optimizer  Trained the model on 127 tuples and attained an average accuracy of 90% on 53 test tuples OTHER PROJECTS House Price Prediction| Dr. Dhiren Bahuguna, IIT Kanpur Aug’18-Nov’18  Developed a Multi linear regression model using Ordinary least square method to predict ‘House Price’ using House Pricing dataset  Performed test of multi collinearity, model adequacy, outlier, variable inflation, and variable selection to find the best fit model Fourier Spectral Methods | Course Project, Dr. Pravir Dutt, IIT Kanpur Jan ’19-Mar’19  Applied Fourier spectral to solve Naiver stokes equation on two-dimensional flat torus with Crank Nicolson method  Used FFT with initial conditions like smooth, uniform random noise, a combination of two to obtain a converged solution Non-linear regression | Course Project, Dr. Debasis Kundu Jan ’19-Mar’19  Estimated parameters of Compartment Model from the given data using Newton, Prony and Osborne methods  Plotted 3D contour for squared error and proposed cross validation approach for a number of components RELEVANT COURSEWORK AND TECHNICAL SKILLS Data Science: Data Mining | Machine Learning |Database System | Data Structure & Algorithms | Linear Algebra | NLP** | Deep Learning* Statistics: Probability & Statistics| Linear Regression | NonLinear Regression | Bayesian Analysis | Statistical Simulation |Stochastic Process Languages: C | C++ | Python | Matlab | R | MySQL | ML frameworks & Utilities: Keras | Tensorflow | Scikit-learn | Latex | Git | Linux *Coursera |**Audit POSITIONS OF RESPONSIBILITY Institute Sports Events | Manager, UDGHOSH Feb’17-Oct’17 Leadership  Led a 2-tier team of over 50+ members in coordination with officials and referees for successful conduction  Introduced a website to maintain database of referees and player’s profiles including Rules, Schedule and Results Initiatives  Facilitated conduction of 5 new formal events, namely handball, kabaddi, skating, sports quiz, aquatics  Planned and organized a Marathon for 11-25 age group in Kanpur which saw over 2000+ footfalls  Successfully conducted all the events with a positive feedback from participant and referees of other institute Impact  Contacted several state and national sports federations, and negotiated them to sponsor the events Institute Cultural Events | Principal Security Officer, ANTRAGNI May’17-Oct’17  Worked in a 3-tier team of 5 PSO, 10 SSO and 30 SO to ensure full security and smooth conduction of all the events   Prepared security plans for crowded places and managed a crowd of over 5-6k people including IITK junta, participants, professors etc Member(elected) | Hall Executive Committee(HEC), Hall 9 Aug’19  As a HEC member, contributed in management of basic hall facilities for uninterrupted, comfortable stay of hall residents EXTRA CURRICULAR     Coordinator, Inter IIT’16: Introduced several new Informal events and prepared a plan of action for smooth conduction of all the events  Volunteered for Prayas to provide basic education of 50+ marginalized kids around IIT Kanpur campus Designed structures based on World War II theme in Galaxy, Inter Hall competition and won first prize Hobbies: cooking, photography, watching movies and TV series, playing indoor games ADDRESS: S/0 – Krishana Dev Chaudhary, Kushdihra , Paliganj , Patna , Pin code – 804426 , Bihar Mob-/-