Predicted Big 5 personality traits using Random Forest and xGboost classifier on eye trackingdataset. Accomplished results better than the state-of-the-art model and currently WIP to publish in a journal.

Extracted the tweets via Twitter scrapping and analyzed them to identify and comprehend the topics of sociological ideas discussed in social media using topic modeling

- Preprocessed the COVID genomic data using the graphical technique to build the dataset with protein spike sequence (k- mers) as predictors to develop a novel technique to predict the virus strain using a tree-based model- Received complete scholarship for the semester

Designed a machine learning pipeline to predict the number of opportunities and theirvalues created by a marketing campaign using Random Forest regressor to help marketer prioritize their campaign

• Designed an exhaustive recommendation platform that classified the customers into different segments based on their current state of journey. The journeys were developed to personalize the marketing campaign. ROI per customer grew by 25%• Predicted the potential churning customers using XGBoost and targeted them with their preferred offers type. (Recall 90%, AUC = 0.87) Generating additional revenue of RM 1.3M• Identified the major pain point areas of VIP customers by understanding the sentiments of thereviews using Latent Dirichlet Allocation (LDA) which resulted in an increase of customer retention by 18%• Forecasted the daily footfall for the world’s largest hotel and optimized the price of the rooms for each segment using ARIMA, LSTM and Prophet• Based on demographic and behavioral parameters clustered the 6.5 million casino customers, to create the best marketing offers and promotion for each persona.• Build a recommendation system using Lasso regression to understand the customer preference towards different promotional mail and then target the customer based on the email features which were important. CTR of email response increased by 25%• Clustering Tool - Created a tool for cluster generation (Algorithm embedded - k-means, k-modes, GMM). Feature of the tool: Data preprocessing & exploration, Data Visualisation, PCA. Reduced the effort of an Analyst by a significant amount to perform the above task.

• Led a team of 8 analysts to provide analytical support and develop a marketing strategy to our E-commerce based clients • Worked on Market Basket Analysis, using Apriori Algorithm and cosine similarity, to predict which combinations of products have a higher association • Predicted the induced effect of an SMS campaign on the Organic and Paid Search channel, using the Linear models. Optimized the campaigns by reducing spend on certain keywords. (25% increase in ROI)• Collaborate with cross-channel teams regarding the budgeting of various campaigns and provide a data-driven decision by measuring the performance of the KPIs on a daily basis • Created a framework to combine online and offline data to build automated reports that brought down the effort from 20 hours per month to 30 minutes.• Worked on multiple channels like Paid Search, Display, Shopping Ads, Facebook and have a deep understanding of effect of each channel in the customer decision

• Analyze web traffic using clickstream data from online tools such as Omniture Site Catalyst, Discover, and Google Analytics.• Mapped the customer journey and optimized the website by removing redundant pages using Markov Chain rules • Handled site optimization projects that include Customer Journey, Micro-segmentation and Drop Off Analysis• Designed a complex SQL system to store and process the data in a centralized way• Understand client requirements and communicate KPI reports and dashboards over client calls/discussions.