Jacques Diambra-Odi
Data Scientist
Self-motivated, curious, diligent, and goal-oriented Data Scientist eager to apply my Python skills to solve business problems. With a background in science, I bring over 8 years of analytical thinking, data analysis, attention to detail and problem solving to the table.
View My LinkedIn Profile || View My Github Profile
Portfolio
Data Science
Featured Projects
Ice Retail
Description
We conducted EDA to determine a profitable strategy in North America, which entails selling the next big Call of Duty or GTA game on the PS4.
Sure Tomorrow Insurance
Description
We provided a very accurate model to predict whether a customer will, or will not receive insurance benefits.
Sweet Lift Taxi Time Series Forecast
Description
We provided a model for Sweet Lift Taxi to predict the number of orders of the next hour, allowing their drivers to anticipate times of high demand.
US Vehicle Sales Web Application
Github Repository || Web Application
Description
This web application displays interactive visualizations for used car sales advertisements.
Spotify EDA Web Application
Github Repository || Web Application
Description
We identified trends in Spotify users and songs, as well as the relationship between music awards and popular culture with the most popular artists on the platform.
Other Projects
Beta Bank Churn Predictor
- Description
In an attempt to define the most cost effective solution to retain customers, we provided Beta Bank with a model to predict whether a member would churn.
Rusty Bargain Car Value Predictions
- Description
We developed a regression model for Rusty Bargain’s mobile app that quickly appraises the value of a car, while also considering the importance of quality in the prediction.
Film Junky Union Sentiment Analysis
- Description
We used NLP to complete the objective of creating a model that could accurately predict negative film reviews, and acheived and F1 score of 0.85.
Interconnect Telecom Churn
- Description
We used boosting classification models to forcast churn of clients, and acheived an AUC ROC score 25% better than our target.
Yandex-Music EDA
- Description
We used data provided by Yandex.music to test hypotheses on user behavior, and preferences in the cities of Springfield and Shelbyville.
Instacart EDA
- Description
Insights gathered on Instacart customers indicated the number of orders placed depended on variables such as time of the day, day of the week, and the time since the customer last placed an order.
Megaline-Plus Marketing Strategy
- Description
As a means to further increase revenue, we determined optimal capital allocation from the marketing budget to Megaline’s more profitable Ultimate Plan.
Megaline Plus Plan Recommendations
- Description
We analyzed subscriber behavior on a legacy plan, to accurately recommended one of Megaline’s newer plans: Smart or Ultra.
Zuber Ride Share Passenger Preferences
Notebook || Web Notebook1 Web Notebook2 || Map1 Map2
- Description
We conducted EDA to understand passenger preferences and the impact of external factors on ride frequency.
Oily Giant Profitable Well Prediction
- Description
We are given oil well parameters in three distinct regions, upon which we will use to create our model to predict the volume of reserves in the new wells, and the region with the highest total profit.
Zyfra Gold Recovery Prediction
- Description
The task was to build a model to predict the amount of gold recovered from gold ore by optimizing production and eliminating unprofitable parameters.