Mail ID: email@example.com
Seeking data scientist/Analytics position with Machine Learning, NLP, Image Processing, Cognitive modeling, developing and optimizing statistical models, Prediction, forecasting, and statistical analysis with proven records in developing state-of-art model along with knowledge of Statistical Analysis and majorly in Client facing role with capability of building SPEED POC.
University/ State Board
Ph.D Computer Science
Domain: Machine Learning & Statistical Analysis
(Prof. V.Ramalingam IIT Delhi)
M.E Mobile and pervasive computing
Velammal Engineering college, Chennai
B.E. Information & technology
Annai Theresa College of Engineering, Tamil Nadu
July 2016 – Till Date
IBM LABS (ISL) – Data Scientist
2015 October - June 2016
Accenture R&D– Analytics Advisory (Team Lead)
MultiGlobal Verus Solution – Data Analyst
2010 - 2011
TIFAC CORE Pervasive Computing Technologies under Department of Science and Technology (DST), Central Govt. of India, New Delhi
Doctoral Research in Machine Learning and Statistical Analysis (2012 - 2015)
Project 1: Cognitive Embodied Conversational Agent based on Natural Language Processing (2012 - 2014)
Worked Two year in implementing virtual agent with information management, Machine learning and engaged with dialogues based on user queries.
Unlike traditional models of machine learning such as spending enormous effort to train huge corpus of data set each and every time a novel learning SVM is implemented for continuous instant learning without scraping old/pretrained model. Included substantial text mining and extraction using a variety of existing Machine learning tools, Python: spacy Scrapy, ScikitLearn, keras, tensorflow, pybrain to compare performance and used R for statistical analysis.
[Received best Machine Learning Application Award in 2014 Annamalai University]
Project 2: Pervasive Health Care Recommendation Engine using Machine Learning Techniques (2014 - 2015)
One year and nine months involved in developing physical therapy recommendation model using Microsoft Kinect 2.5D camera which will be acquitting 2D and Depth images. ML algorithm will be tracking the human skeleton and in turn human skeletal angular movements are provided as a feature to ML algorithm which will classify how good user/patient have performed their physical therapy and result will be produced in form of precise summary for easy understanding by user/patient and also evaluated the statistical value of the solution and make a recommendation out of findings. Included development of a LV Merge to assist simulation practices using C++ and Python programming. (https://sites.google.com/site/rajaucse1)
Implemented deep learning methods and weak classifiers such as random forest tree for real time analysis.
TIFAC- CORE Pervasive Computing Technologies (2010 -2011)
TITLE: Research Intern
One year as developer in TIFIC CORE – Using challenging Indian Central Government Biometric data set (UID – Aadhar Approx. 200 million records) preliminary analysis was conducted to choose best machine learning algorithm for recognizing identities based on query data along with ApacheTM hadoop.
Implemented various machine learning algorithms such as support vector machine, Auto Associative Neural Network, Genetic algorithm, HMM, EMM, etc., and clustering algorithms like fuzzy, k means, c means, PCA, ICA, etc.
Conducted statistical analysis using z -test, t – test, Cross fold validation, Confusion matrix and ROC to measure performance.
Finally founded, a probabilistic modeling approach using Bayesian classifier with PCA based clustering yield high accuracy ratio.
Multigoal Verus solutions
I)Data Cleansing using python- pandas and numpy Regex
II)Involved in mining suspicious IP from available data set using machine learning for crypteianetworks using logistic regression
III)Finding outliers using Unsupervised learning and box plot
IV)Integrating C++ modules, multithreading, sync and async communication and creating rest api with Microsoft SQL Server as backend database
V)Microsoft azure – Chat bot, Cognitive services
VI)Building Docker containers and deploying into kubernetes, send logs to kafka
VII)Handling Distributed Cloud Object Storage and Querying/Accessing data from various Buckets.
Developing micro services and Continuous integration (Devops)
Accenture R&D – Data Scientist
I)Developed model for prioritizing Resume using python and R
II)RESUME parsing using NLP and prioritizing based on education and skill sets using Python – Used regex and semantic model for Data Extraction and prioritized based on average weighting.
III)Medical Document classification and recommendation using Cognitive Modeling by extracting drug details, symptoms, reaction of drug, patient age, gender, etc using Python & NoSql
IV)Email classification – Extracting data from To, Subject, entities from body of the mail and extracting meaningful using python, NLTK, SPACY, Knowledge graph and customized POS tagging.
V)Developed BI visualization for e-commerce products.
IBM LABS – Data Scientist
I)Implemented Customer Segmentation using Churn Analysis and recommendation for retaining customers.
II)Extracted Campaign insights using exploratory analytics using Regression
III)Performed hypothesis testing and feature selection for Regression
IV)Performed Cognitive 360 deg. insight reporting model for various data using customized Natural Language Generation.
V)Developed Insight extraction and recommendation for various campaign data using Risk analytics and Fraud Detection
VI)Developed Natural Language Understanding (NLU) – For summarizing the content using entities and blob text and Bert.
VII)Implemented Document Classification using CNN for Watson Education.
VIII)Architected and developed Content Linker concept end to end. For content matching and retrival – Improved accuracy, speed and added NOSQL Mongo DB with concept of Semantic similarity and TFIDF for Watson Education using GENSIM.
IX)Website Crawling and did Graph QL analytics for extracting distribution and frequency of a topic available in social media content.
X)Applied OCR for retrieving text from image document and optimized accuracy using image optimization and spellcheck.
XI) Tried POC in building Sequence Vector Model using LSTM for generating Keywords.
XII) Customized Topic Extraction from the feedback and visualizing
XIII) Video Analytics – Searching meta data inside video using semantic and syntactic keywords using Glove.
XIV)Developed Building insights based on occupancy and energy consumption (Watson IOT)
XV)Performed space planner optimization using headcount and space availability (Watson IOT)
XVI) Research and prototyped various Anomaly detection technique for time series data.
XVII)Did research in building insights using ARIMA and ARIMAX – performed time series modeling.
XVIII)Building docker container image and deploying into kubernetes.
XIX)Re written existing analytics querry and code to micro services.
XX)Worked in statistical optimization for various distribution of data using Scipy (COBYLA)
Received 3 times Manager choice award.
Received External Eminence Award.
IBM AOT - Tried POC in Augmented reality and got appreciation from US team.
IBM AOT – Completed building Cognitive Chat bot.
Hackathon Call for Code (2018) - Owned best use case prize in ISL (Got granted: $1000)
Hackathon Call for Code (CFC) 2019 – Owned best application in ISL (Got granted: $5000)
Submitted IP and went for search 1- A 360-degree cognitive insight reporting tool
Prototyped and showcased Self flying Drone with human detection using Image processing and machine learning.
Actively participating in AI-a-thon 2019.
Active speaker in colleges – regarding Machine Learning and IBM ML services.
Windows, Red Hat
Programming Languages & Packages :
Watson Discovery Service, Watson Machine Learning, Watson Visualization, Watson visual Recognition, Watson neunets, Watson Deep learning, Java, Python, R programming, SAS, Matlab, Octave, MySQL, NOSQL, MongoDB, IBM DB2, Shell scripting, Ruby, Open CV Open ML, C/C++, C#, Weka-3, NLTK, Open NLP, Scrapy, scipy, ScikitLearn, Deep Learning, pybrain (Open source Machine Learning S/W), JSON, BSON, SQL, OOPS concepts, Image Processing, Data Mining, Text Mining, Microsoft Azure, Bluemix, Watson Studio, pytorch, YOLO, SSD, Solr search, Elastic search, Word2vec.
Machine Learning Algorithms (Neural network, Genetic, Fuzzy, SVM – polynomial, Gaussian, RBF, MLP), XGBoost, Prophet, Natural language processing (NLP), Prediction modeling, Regression, Deep Learning, Data Mining – (Complete, Incomplete, Structured and Un structured data set), Time series analysis, Data warehousing, Data Management, Text mining – GENSIM, Glove, Semantic Similarity, Segmentation, Feature Extraction, Processing Structured and un Structured Database, Testing (ANNOVA, FRIEDMAN) Data Structures, PCA, Clustering, Statistical analysis and modeling – ARIMA model, etc., Cloud computing, Big Data, Graphical visualization (Graphs, Tables and charts), Convolution Neural Network (CNN), FRCNN, RNN, Long Short term memory (LSTM), RNN, Object detection – Darknet, Yolo .
Academic Projects :
Project in M.E
Association rule mining with hashing and pipelining for statically analyzing unstructured data base
This project involves face recognition using Neural Network classifier and compared performance of various machine learning algorithms against FERET face database and fusing multi modal features.
Project in B. E
A comparative performance study of a fine-grain multi-threading model on distributed memory machines
This project presents a comparative study in implementation of threads in distributed system using python.
Completed Java and C# with .net in NIIT Pondicherry during year 2009.
Azure Machine Learning from Microsoft during year 2013.
Completed IBM DB2 course Dallas Technologies during year 2013.
Workshop and publications:
UGC SAP DRS-1 Annamalai University sponsored inter disciplinary research methodology workshop.
National Research Methodology workshop on speech and image processing.
National workshop in pervasing computing technologies.
Conference / Journal
International conference on Soft Computing, Optimized face recognition using 2D wavelet decomposition principle component analysis with singular value decomposition using bayesian classifier.
IEEE conference, Intelectually combined face recognition using curvelet based principle component analysis for feature extraction and bayesian classifier.
International conference on signal processing, Intelligent face recognition using 2d wavelet transformation using SVD.
IEEE pattern analysis and machine intelligence Realtime human detection from cascaded SVM classifier using raster scanning approach (Accepted Waiting for publication).
International Journal of Applied Engineering Research “Novel random forest tree based human action recognition using rgb-d sensor”
International Journal of Applied Engineering Research ”A Novel Fuzzy Rule Based Occluded Facial Expression Recognition System Using Expression Overlapping Technique”
Given lecturers in various Machine learning algorithms (Supervised, Unsupervised, Regression classifier, Clustering, SVM, EM, GMM, HMM, SOM, Neural networks, Genetic algorithm, fuzzy etc.,) and estimation of classifier performance (bagging, bootstrap, cross validation, etc.,) in Annamalai University.
Senior member in IEEE.
In plant training
Undergone implant training in BSNL Chennai during year 2009.
Bangalore - Karnataka
English, Tamil, Kanada
Declaration: I hereby declare that the information is true to the best of my knowledge.
Place: Bangalore RAJKUMAR NATARAJAN