SUMMARY

• Extensive working experience in the field of Data Analytics, Data Science, Machine Learning, Data Mining, Predictive Modeling, Descriptive Analysis, ETL Development and Data Visualization.
• Proficient in Statistical Modeling, Data Mining and Machine Learning Algorithms in Forecasting/ Predictive Analytics such as Linear/Logistic Regression Boosting/Bagging, Ensemble methods, SVMs, Naive Bayes, Random Forests, Ada Boost, Gradient Tree Boosting, XGBoost, Decision Trees (CART), Clustering Algorithms (K - Means, DBScan, Agglomerative), Feature Engineering and Dimensionality Reduction (Principal Component Analysis, Feature Selection/Elimination and other statistical techniques)
• Skilled in Deep Learning Framework: TensorFlow, Keras, Kubernetes, and PyTorch; Familiar with Deep Learning Models like ANN, CNN, RNN and LSTMs.
• Worked with model validation using cross validation, Stratified K-Fold, Hyperparameter Optimization using with Grid and Random search; used Lasso and Ridge Regularization to avoid the overfitting in the model.
• Measured and Tested model performance using AUC-ROC curve and Confusion Matrix and identified accuracy, precision, recall and F1, RMSE, MSE and MAE combination of algorithm parameters specified in a grid, finally we increased accuracy by 5%
• Expertise and knowledge of multiple programming languages like Python, C#, MATLAB along with sound fundamentals of Data Structures and Algorithms.
• Extensively worked on PyCharm and Anaconda Navigator (Jupyter Notebook) with libraries like Numpy, Pandas, Scipy, Scikit-learn, matplotlib, seaborn packages.
• Hands on experience on AWS Cloud Infrastructure like EC2, S3, RDS, Redshift, Data Pipeline, DynamoDB, Lambda, Athena, Kinesis, IoT, IAM, Quicksight, Elastic MapReduce.
• Experience in data mining and bringing structure from data that is semi and unstructured using SQL, SAS, Python, advanced MS Excel (VLOOKUP, Powerpivot, Pivot tables) for supporting Data Migration, Data Cleaning, qualitative and quantitative analysis.
• Highly proficient in using RDBMS like T-SQL and Postgre SQL for developing complex Stored Procedures, Triggers, Tables, Views, User defined Functions, User profiles, Relational Database Models, Data Integrity, SQL joins, indexing and Query Writing and also experienced in non-relational databases like MongoDB 4.0.
• Well versed in ETL tools like SSIS 2016, Informatica Power Center 10.1.1 and Pentaho Integration.
• Strong understanding in Statistics methodologies such as Hypothesis testing, ANOVA, Chi-Square and A/B testing to find the accuracy of model and time series analysis: moving averages, ARIMA.
• Defining KPI's and building predictive models, classification models and communicating results.
• Delivered BI reporting solutions in Reporting services (SSRS), SAS Analytics and MicroStrategy and created Tableau 2020.1 dashboards, visualizations, and performing advanced analytics.
• Involved in SDLC- software development life cycle (Water, Scrum/Agile) of building a Data Warehouse on windows.
• Excellent Knowledge on Hadoop Ecosystem such as Hive, Spark 2.3.4.
• Extensive experience with version control tool Git and GitHub.
• Proficient in Microsoft Project, Microsoft Office, ProjectLibre, JIRA, Docker.
• Proactive and sincere motivated team player with exceptional problem-solving, analytical, interpersonal, innovative skills and excellent attention to details.

TECHNICAL SKILLS

Languages \ Big Data Tools:: Python 3.7/3.6.4/3.5.2/3.3 , SAS 9.4, SQL, C# \ Spark 2.3.4, (PySpark, SparkSql) \ T-SQL, HiveQL\

BI Tools \ Operating Systems: \: Tableau 9.4/9.2/2020.1 , MicroStrategy, \ Windows 10/8/7, Mac OS, UNIX, Linux\ MS Office (Word/Excel/PowerPoint/Visio) \ SAS Visual Analytics\SSRS\QlikView

Database \ Report/Document Tools: \: Oracle11g, MS SQL Server 12.0/14.0/15.0/ \ MS Office 2016, MS Project, Outlook, Excel, \ MySQL5.7.6\ 5.7.11\8.0.18\ PostgreSQL11\ Word, PowerPoint\ Mongo DB 4.0\

Machine Learning/Deep Learning \ Packages: \: Regression models, Naive Bayes, Decision trees, \ Numpy, Pandas, Scipy, Seaborn, Matplotlib, \ Random Forests, Ada Boost, XG Boost, SVM, \ Flask, Plotly, Pytorch Keras, Scikit-learnKNN, Bagging, Gradient Boosting, LDA, \ TensorFlow, Kubernetes\ K-means, Neural Networks, CNN, RNN\

ETL Tools \ Infrastructure: \: Informatica Power Center 10.1.1, \ AWS S3, EC2, RDS, Redshift, DynamoDB SQL Server Integration Services 2016 Lambda, Kinesis, Athena, Google Colab, Docker

Tools: \: Git, GitHub, ProjectLibre, JIRA\

PROFESSIONAL EXPERIENCE

Confidential

Data Scientist

Responsibilities:

• Created Kinesis Data Firehose delivery stream to transfer the data to AWS S3 bucket.
• Configured the Lambda function to periodically import the data into Amazon EMR from AWS S3 and then further loaded data to Redshift.
• Conducted Exploratory Data Analysis (EDA) on the data using Matplotlib and Plotly in Python using to visualize and gather insights from the data, test prior assumptions and identify the missing values and important variables.
• Performed Feature Engineering for data cleansing such as detecting outliers, missing value and interpreting variables, convert ed the categorical features to numerical features.
• Used Plotly in Python to analyze the data distribution of the sale price and its correlation with other numerical features.
• Used PCA for Dimensionality Reduction and to find the strong patterns in a dataset.
• Built various models like linear regression, decision tree, SVM, Random forest Regression, Gradient Boositng, XGBoost, LightGBM to find out the best fit model by achieving lowest prediction error.
• Used GridSearch to tune hyperparameters and evaluate a model for each combination of algorithm parameters specified in a grid, finally we increased accuracy by 5%
• Model testing is done using evaluation metrices such as Root-Mean-Squared-Error (RMSE), MSE (Mean Squared Error), MAE (Mean Absolute Error) between the log of the SalePrice predicted by the model, and the log of the actual SalePrice.
• Created views and visualizations in Quicksight and Tableau Desktop for monthly/ quarterly KPI data using filters and actions and presented the predicted monetary value of the houses. Used GitHub for version control.

Environment: Python 3.x(Numpy, Pandas, Scikit-learn, pykalman, pydlm), AWS S3, Redshift, Lambda, Quicksight, MS SQL Server 15.0, Tableau 2020.1, TensorFlow, GitHub

Confidential, Denver, CO

Data Scientist/Data Analyst

Responsibilities:

• Performed the extensive Exploratory data analysis like descriptive statistics, histograms, boxplots, correlation of values, ANOVA table, confidence intervals, Normality curve, residuals to select the best model
• Conducted the data cleansing and preprocessing, applied moving average to fill the missing values, like putting dummy values (Boolean numbers for few columns), using library and string functions and eliminating the missing values.
• Visualize and analyze the data using matplotlib, seaborn to find out the key universities and colleges with low ranking, factors affecting ranking and provided recommendations based on the results.
• Built multiple linear regression models to predict the ranking of the universities for the year 2019, calculated error percentage between training and test data.
• Performed k-fold Cross validation, error metrics to scrutinize the predictive ability of model and optimized the model for the better results.
• Designed presentation ready dashboard using Tableau to monitor the Key Performance Indicators (KPI) and determine the top-ranking universities in the Colorado region.

Environment: Python3.6.4 (Numpy, Pandas, Scikit-learn, matplotlib seaborn), MS Excel 2016, Tableau 2018.1, GitHub