Navdeep

Data Analytics/Scientist

Analytical Tools:

R, Microsoft Excel, WEKA, MATLAB, Minitab

BI and Visualization:

Tablaeu, SAP Lumira, QlikView

Languages:

R, SQL, Python, FORTRAN, C

Databases:

RDBMS, Oracle 11g, MySQL

Data Warehouse Tools:

Data Profiling, Mapping and cleansing, OLAP, OLTP, Conceptual, Physical and Logical Data Modeling

Geographical Information System (GIS) and Remote Sensing (RS):

ArcGIS Desktop, ArcGIS Server, Python for ArcGIS, Web AppBuilder for ArcGIS, ArcGIS Online, Quantum GIS, Erdas Imagine, Trimble eCognition, Global Positioning System (GPS), Trimble GPS-GEOXM, Google Earth. Knowledge of ArcObjects, ArcGIS Data Reviewer, Geodatabase Domains and Subtypes.

Other tools:

MS-Office, Gretl, MathType, LaTeX, PV-Wave, NetLogo, Generic Mapping Tool (GMT), PSI-Plot, Petrel, Seismograph (UL 408)

Analytical Concepts:

Univariate & Multivariate Analysis, Linear & Multiple regression, Logistic regression, Principal Component Analysis (PCA), Factor Analysis, Decision tree (CART, C5.0, CHAID), Segmentation, Cluster analysis, Time series analysis, Text mining and Spatial statistical analysis

Domains of Interest:

Energy, Geospatial, Insurance

Operating Systems:

Sun Solaris, Linux, Windows 10/8/7/XP, MacOS

Johnson Controls, Inc., Syosset, NY                                                                                       December 2015 – present

Data Scientist

In the project “Smart Grid Analytics and Energy Price Forecasting”, I have been contributing towards the development of smart energy analytics solution portfolio for its utilities and industrial clients to integrate and manage Smart Grid and Renewable Energy infrastructure. Such systems help the clients to reduce the carbon footprint, obtain real-time visibility of demand and supply, and optimize the energy generation, distribution in a cost effective fashion. The overall aim of the project is to collect data in (near) real time and integrate with the existing ERP backend system to facilitate generation planning and control, demand flexibility, dynamic pricing, demand forecasting and, maximize the utilization of renewable sources by predicting overall consumption and supply patterns.

Job Description:

·   Develop BI and analytics reports on demand and supply optimization trends and KPIs.

·   Extracted test data from proto-type equipment, weather data from the National Oceanic and Atmospheric Administration and solar radiation data from the National Aeronautics and Space Administration (NASA).

·   Developed SQL queries for extracting utilities usage data to lower environment for data analysis. 

·   Explored Hive QL for summarizing, querying and analyzing data stored in HDFS.

·   Explanatory data analysis and data health check.

·   Performed outlier detection, imputation for missing data, cleaning, transforming, and normalization.

·   Coordinated with domain consultants and technical team for product development and customization.

·   Coordinated with client SME’s to define the blueprint for data and service integration.

·   Performed literature review and research to identify historical patterns and methods applied.

·   Clustering to obtain similar type of data based on location and availability of alternate energy source.

·   Developed various non-linear regression models to study the relationship between variables.

·   Applied ARIMA method to forecast renewable energy usage.

·   Built a predictive model to quantify structural reduction in net load as a function of the presence of solar PV, house and household characteristics and weather for a particular zone.

·   Develop modules for price performance, asset performance and alerts.

·   Aggregate relevant information and developed tableau dashboard to present the results.

Environment: R 3.2.3, Microsoft Excel 2013, Tableau 9.2.2, MySQL 5.5.48, Hadoop, Hive

Air Products and Chemicals, Inc., Allentown, PA                                                        August 2015 – November 2015

Geospatial Analyst (GIS)

Computational Modeling Center (CMC) group, hired me to develop geospatial models and perform object based image analysis for making informed business decisions. Involved in spatial data collection and munging, creating geodatabases, and performing necessary analytics. Generated progress reports and manuals to communicate with team members and provide guidelines for efficient spatial analysis. The location specific results and reports were used by the project manager and department head for making business decisions.

Job Description:

·   Spatial data preparation including collecting shapefiles, grid generation, and file creation.

·   Summary statistics using Analysis tools.

·   Location data validation using basemap in ArcGIS.

·   Developed a conceptual framework for location based spatial analysis.

·   Identified key variables (shapefiles) for the analysis.

·   Outlier detection and cleaning using proximity analysis tools of ArcGIS.

·   Merged, modified data and extracted attributes using data management tools.

·   Analyzed the area and identified locations for detailed investigation.

·   Developed python scripts for automated file modification and management.

·   Object based image analysis to identify desired objects using Trimble eCognition software.

·   Validation of results using KML and KMZ files in ArcGIS and Google Earth.

·   Geocoding of results for detailed location information using ArcGIS.

·   Exported the results in different file formats for further validation and presentation.

·   Streamlined the project by building geoprocessing workflows using model builder in ArcGIS.

·   Increased efficiency by automating the project using python scripting in arcpy.

·   Performed geospatial analysis and image processing to identify desired objects in each given locations.

Environment: ArcGIS Desktop 10.3, Python for ArcGIS, eCognition 9.1, MS Excel 2013, Google Earth 7.1

Advanced Material Analytics, LLC, Binghamton, NY                                                            June 2014 – August  2015

Research Scientist (Data Analytics)

The goal of the project was to develop commercial data acquisition system for material characterization. In this I have been leveraging my expertise in geophysics and data analysis to integrate complex statistical tools with hardware systems. These equipment are used for testing “materials/products” for characteristic properties like porosity, permeability & pore-size measurements.

Job Description:

·   Supervised the research and offshore software team through equipment and software development cycle.

·   Developed MATLAB programs to identify the proto-type equipment parameters and component selection.

·   Efficient design of sample chamber and equipment body using PTC Creo.

·   Obtained multiple dataset from the proto-type equipment for data analysis.

·   Performed data health check and summary statistics.

·   Applied regression and correlation methods to illustrate the accuracy, repeatability & reproducibility of the equipment.

·   Explored automated report generation using python scripting.

·   Developed SQL queries for client information management and update.

·   Explored and created tableau dashboard to present the results.

·   Offered recommendation on the feasibility of the project, based on simulation (of the pycnometer method) and deviations from competitor’s data.

Environment: MS Excel 2010, R 3.1.2, MySQL 5.5.32, Tableau 8.3.4, MATLAB 7.12, PTC Creo 3.0

Dept. of Geological Sciences, Binghamton University, Binghamton, NY                              June 2008 – May 2014

Research Assistant (Data Analytics)

The main research was centered on the data processing and modeling of the 12 years of TOPEX satellite data to compute the changes in sea-level due to atmospheric pressure variations, and the effect of the changing sea-level on Earth’s rotation. As a research assistant, I was asked to perform multiple small projects related to soil moisture analysis, ecological habitat, and transportation analysis, using R, Python and ArcGIS.

Project 1: The goal of this project was to model the spatial and temporal variability of the oceans due to atmospheric pressure using TOPEX satellite data. The secondary goal was to calculate the effect on the Earth’s rotation due to ocean-atmosphere interactions.

Job Description:

·   Obtained TOPEX satellite data from NASA-JPL in netCDF format and extracted them to obtain Big Data of 216 gigabyte (~300 million data records).

·   Pre-processed the data including data cleaning, missing data imputation, transformation, and normalization.

·   Performed detailed literature review and research to understand the problem and identify popular methods applied in this domain.

·   Corrected the complex-numbered sea level time series for effects due to tides and winds; applied conservation of mass.

·   Obtained the daily atmospheric pressure data from Atmospheric and Environmental Research, Inc. and sampled the data for time period of 3 days.

·   Performed 3-dimensional spatial and temporal interpolation in FORTRAN for missing data points.

·   Reduced the data using weighted Gaussian averaging and applied rules for coastal boundaries.

·   Fitted a sinusoidal regression model with a trend line to isolate the effects due to seasonality.

·   Applied Wiener filtering – an advanced autocorrelation and cross-correlation method – to calculate the change in sea level due to atmospheric pressure variations.

·   Selected the optimal filter length by reducing the root mean square error between input and output.

·   Developed a novel data based model for the ocean-atmosphere interactions and calculated its effect on change in Earth’s rotation speed and orientation.

·   Presented the results at multiple conferences and submitted a journal article based on this method.

Environment: MS Excel 2010, Fortran 90, PSI-Plot 8.8, Python 2.7

Project 2: The goal of this project was to simulate the ocean-atmosphere interactions and establish the spatial and temporal variability in the sea-level change due to atmospheric pressure fluctuations. The main aim was to check the validity of the static “inverted barometer” model (widely used in this domain).

Job Description:

·   Detailed literature review for developing the ocean-atmospheric interaction model.

·   Developed a local as well as global static model in FORTRAN to observe the response to pressure.

·   Obtained and modified mathematical equations for the development of a “true” ocean-atmosphere model.

·   Developed Green’s function based spatio-temporal model and simulated using unit amplitude atmospheric pressure on the oceanic surface.

·   Applied spherical harmonics for global representation of the oceanic response to atmospheric pressure.

·   Illustrated the results by creating global plots using PSI-Plot software.

·   Established the significance of a dynamic model compared to the static model.

·   Presented the results at American Geophysical Union conference (biggest conference in Geoscience domain) and published a journal article based on the project.

Environment: MS Excel 2010, C, Fortran 90, PSI-Plot 8.8

Project 3: This project involved the fish habitat mapping using fish-availability data, Stream-flow analysis (USGS-NHD datasets) and Geomorphological datasets. This study was aimed to understand the relationship between geomorphological variables with the ecological datasets

Job Description:

·   Extracted data from various sources, cleaned, transformed, and normalized them and created ready to process data files.

·   Applied ANOVA for testing the differences among years.

·   Developed a linear regression model to establish a relation between a fish species and river water velocity.

·   Performed K-means clustering based on location as well as species.

·   Evaluated different number of clusters and their significance using R.

·   Established relationship between different fish species and their habitat.

Environment: MS Excel 2007, R 2.15.3

Project 4: This project involved the calculation of soil moisture content of an area from soil texture data and land-surface model data. The soil water content of the given region were calculated, then compared the results with the model data and explore different methods of handling missing soil layer data.

Job Description:

·   Summary statistics of the sensor data.

·   Performed data cleaning and missing data imputation using R.

·   Reviewed different literature to establish relationship between soil water potential and soil water content.

·   Calculated the soil moisture content based on the data.

·   Compared the results with the land surface model output using correlation and regression.

·   Explored different missing data imputation methods and interpolation techniques to reduce the error between the observed data and land surface model.

·   Calculated critical soil moisture zones and mapped using ArcGIS.

Environment: MS Excel 2007, ArcGIS 10.1, R 2.14.2

Project 5: This project involved the analysis of NY taxi trip data. The data contains taxi trips with starting and ending location, duration, fare and payment methods used. The main aim was to analyse revenue generated per driver, shortest and longest distance travelled between 2 points and preferred payment methods by location.

Job Description:

·   Performed data munging of the 5 gigabyte of taxi trip data.

·   Applied imputation for missing data and created subset of data based on starting location.

·   Calculated mean revenue generated by each driver based on location.

·   Identified the preferred payment method and average tips (based on the base fare) for trips starting at JFK.

Environment: MS Excel 2007, R 2.14.2, Python 2.7

Dept. of Geology & Geophysics, IIT-Kharagpur, Kharagpur, India                                  August 2006 – May 2008

Research Analyst

This project was to study the gravity anomaly of the Orissa region of India and predict the tectonic activity based on past gravity data. Performed data cleaning and explanatory data analysis. Applied correction to the observed gravity data. Then applied forward modeling of corrected gravity data to model the structure of the basin using MATLAB.

Environment: C, MATLAB 6.5, MS Excel 2003, Generic Mapping Tool

National Institute of Oceanography (NIO), Goa, India                                                             June 2007 – July 2007                                                                                                                                       

Data Analytics Trainee