Projects

Solving business problems with machine learning A short project to showcase my knowledge of solving business problems using data science and machine learning techniques. I used a transnational data set which contains all the transactions occurring between 01/12/2010 and 09/12/2011 for a UK-based and registered non-store online retail. The company mainly sells unique all-occasion gifts. I gleaened the following insights from the data: Business Metrics: Monthly Revenue, Growth Rate, Active Orders, Average Revenue per Order, Revenue from first time and returning customers, etc. Consumer Metrics: Cohort analysis, Retention rate, Churn rate, Cohort-based retention rate Customer Journey: From acquisition to activation, engagement and retention Customer Growth: Market Attribution models, Channel optimisation (First-touch, Last-touch, Markov Chain) Customer Segmentation: Clustering and Segmentation (using RFM Recency, Frequency and Monetary Value)

Is the movie industry dying? I provide recommendation about the type of movies a film production company would have to do if the box takings and profit have to be maximized. As an example, I provide answers to the following questions that the company could have: In which genre should the film company focus on? How important is to produce a good film (high IMDB rating) to get lots of takings? Is it better to produce a film for a general audience or is it better to produce a film targeted to a particular segment? Does Higher Budget translate to more Revenue for Animation Movies? Is there an actor/actress that guarantees the success of the film? Or a director?

Customer spend, Satisfaction and Segmentation using Machine learning This study proposes ways drive customer spending using information from customers and applying machine learning methods without spending money on advertising and marketing. The dataset for this study was generously provided by Olist, the largest department store in Brazilian marketplaces. In this study, first, customer segmentation is done by grouping customers into groups based on similar behaviours (such as purchasing, demographics, etc.). This is important because customers have different needs and, as Olist’s customer base grows, it becomes difficult understanding the individual needs of each one of them. Hence, with segmentation I identified customers’ needs and differences in groups and act on them. Second, I identified satisfied and dissatisfied customers using satisfaction ratings, then I developed a model that predicts if a customer will be satisfied or not based on their purchase patterns and demographics Finally, I developed a model for predicting customer spend (that is, the total amount a consumer will spend on each transaction).