Process Analyst Data Scientist

Iftekhairul Islam

B ********************@*****.***

Æ +1 (224) - 595 - 7131 https://www.linkedin.com/in/iftekhairul-islam/

***** ********* ** ******, ** 75035 https://iftekhairul-islam.github.io/Iftekhairul-Islam/ Summary

Economist and Data Scientist with good grasp on Statistics and GIS. 6 years of experience in team leading and managing projects for financial software development, machine learning, spatial analysis, and data science solutions for business problems. 5 years of research experience with grant/proposal writing and managing. Hands-on experience with big data analysis and implementing in cloud platforms (e.g. AWS, Azure). Skills

Natural Language Processing / Machine Learning / Deep Learning / Data Science Supervised/unsupervised machine learning algorithms, features selection and dimension reduction, feature engineering, Deep Learning including CNN, RNN, LSTM and embedding models (e.g., BERT), Deep Generative Models, Generative Adversarial Networks (GAN), Progressive Growing of GANs, Deep Reinforcement Learning, Matrix factorization and Collaborative filtering. Optimization / Statistics / Mathematics

Linear and Logistic regression, integer and mixed integer programming, convex/non-convex optimization, parametric/non-parametric models, statistical inference, sampling, Frequentist and Bayesian statistics, hypothesis testing. KNN, LDA/QDA, Ridge/Lasso, Bootstrap/Cross Validation, Clustering, Decision Tree, Bagging/Boosting/Random Forest, SVC/SVM, PCA, PAFA, Spatial Analysis, Areal Interpolation, Krigging, Spatial Autoregression etc.

Languages/ DB/ Technologies

Python, R, STATA, Scikit-learn/Keras, Caret, Tidyverse, SparkR, Spark NLP, PyTorch, Apache Spark, Apache Hadoop, SQL Server, AZURE Studio, AWS SageMaker, Tableau, ArcGIS, QGIS, unstructured/semi-structured/structured data handling, data crawling, cleaning, wrangling, warehousing. Projects and Grants

Written and led projects and grants including a World Bank funded (USD 236,714, from 2011 to 2013) project, experienced in dealing with small/large scale industrial and government funded software/web applications for different clients including Bangladesh Army. Education

Doctor of Philosophy Public Policy and Political Economy The University of Texas at Dallas Aug. 2016 – Aug. 2021 Dissertation: 4IR and Technological Unemployment: A State-level Investigation Master of Science Geographic Information Science

The University of Texas at Dallas Aug. 2018 – Aug. 2021 Master of Science Finance

The University of Texas at Dallas Jan. 2011 – Dec. 2011 Bachelor of Science Mechanical Engineering

Bangladesh University of Engineering and Technology Jan. 2001 – May 2006 Work Experience

Research Scientist Nov. 2021 – Present

Public Policy Research Institute Texas A & M University, College Station, TX

• Applied Glmnet algorithm to predict ’Bond Failure’ probability of the defendants form Harris County, TX.

• Built a predictive model applying SVM to classify defendants from the City of Houston who are prone to re-offend.

• Developing a CNN model for Cancer Prevention and Research Institute of Texas (CPRIT) to predict spread of cancer in the smoke-free cities in Texas considering spatial and temporal determinants.

• Providing consultation to Texas Workforce Commission (TWC) to model effectiveness of re-skilling and re-training programs for unemployed Texans.

Data Scientist May. 2020 – Nov. 2021

Software Development Division Tasacom Technologies Inc.

• Implemented predictive models using Machine Learning algorithms such as- Linear regression, Logistic Regression, Random Forest, and applied concepts of R-squared, R.M.S.E, P-value, and F test, in the evaluation stage such as to predict client’s financial performance and key variables that influence more towards the profitability of the company.

• Worked on a Deep Learning project with Recurrent Neural Network to determine prospective diagnosis for a certain patient.

• Performed feature scaling, feature engineering and statistical modeling and write complex SQL queries and stored procedure to run daily job for Data Warehouse and the Power BI Dashboard.

• Performed in-depth statistical analysis and data mining methods using Python and used Azure Machine Learning to set up the experiments and create web services for the predictive analytics.

• Coordinated with the backend and the UI teams to generate real-time analytics dashboards for clients.

• Analyzed election donation data to detect racial, ethnic, and religion impact on donation pattern. Business Process Analyst Mar. 2012 – Aug. 2016

Software Development Division Electronic Business Solutions Ltd.

• Developed Power BI and Tableau dashboards for efficiently visualizing cost and budget data for optimized allocation of fund.

• Applied machine learning algorithms for predicting the efficacy of digital marketing employing search query data using google analytics that helped my clients revamping revenue by 20%.

• Used advanced data analytics using QGIS, SAS and SQL for analyzing the growth of e-commerce in Bangladesh and detecting the pain points that helped clients in taking focused business development decision.

• Got involved in client-facing consulting engagements serving clients in Finance, IT, and RMG industry to take actionable and cost-efficient strategy.

• Assisted senior management to drive the strategy execution through designing new business models; streamlining and automating existing business processes.

Data Projects

• Investigate Precinct-level election data (Texas) and ACS data to pinpoint dormant Hispanic voters for ‘geo-coded’ promotion.

• Asses the impact of inaccessibility on labor productivity.

• Analyze O*NET and BLS dataset to predict job vulnerability due to automation and the skill set required for the fourth industrial revolution.

• Assess social mobility and redistribution data to measure federal policy impact on youth employment applying SVAR model using Python.

• Examine impact on jobs due to Covid-19 in inaccessible areas using block-level employment data and ArcGIS Pro.

• Uncover land-use change in the Dallas Metro area using LANDSAT images and ArcMap, producing numerous geo-coded visualizations

Contact this candidate