Rodrigo Marques Portifolio

Here you can find the summary of my main projects as well as their links for more detailed exploration.

---

Understand World Happiness

This kind of study can also be applied to understand which aspects affect customer and/or employee satisfaction in a Business.

• Built a deep analysis on which factos more contribute to the perception of happiness around the world.
• Modeled and Optimized a model to better estimate the score the Countries would get based on each of the attributes (MAE ~ 0.38).
• Tacked questions like: *Is this GDP per capita which makes you happy? *Is this Perception of Corruption about Goverment, which make you sad? *Is this Freedom of Life Choises which makes you happy?

Handwritten Digit Recognition CNN Practice

The MNIST dataset is a well-known dataset for image classification. Tensorflow and Keras also provide MNIST dataset directly through their APIs. For the learning purpose, the MNIST dataset is easy to use and experiment with different machine learning techniques.

• In this notebook, I have covered the necessary steps to approach any Machine Learning Classification Problem.
• Included Image Visualization for better understanding.
• Quick Links to the functions I have used to explore it in depth.
• Basic techniques such as Confusion Matrix, Image Augmentation, etc.
• I have also compared the results of Model without using CNN and CNN.

Time Series Analysis and Forecast of Total Factor Productivity (TFP)

This project consists of an evaluation and comparisson of the Total Factor Productivity for three countries: EUA, Canada and Mexico.
Economists have found that they can explain only a portion of economic growth by the growth of inputs to production, such as the number of hours worked or the amount of capital used. The unexplained (or residual) portion, which presumably reflects advances in production technologies and processes, is conventionally attributed to all of the production factors together and is referred to as Total Factor Productivity (TFP) growth.
The measure of TFP is based on 39 variables that measure relative levels of income, output, inputs and productivity for 167 countries between 1950 and 2011. Productivity growth is often assossiated with new products, tools and processes that reduce costs of estraction and/or transcorming inputs into finished products.
From 90's to 2000's, the TFP showed a global growth trend concentrated in manufacture sactor, and particularly in Information Technology.

Over the long term, TFP growth is limited only by the ability of innovators to develop new technologies, and that a larger population makes possible a larger pool of talent to be devoted to research, and thus opens up more potential for innovation.

Sentiment Analysis for an Mobile App

Situation: The client is a financial sector app that aims to achieve greater visibility and retention of users in an organic way. As this market has grown a lot and is very competitive, he decided to invest in ASO, focusing on his efforts not only in views but also in installations, that is, not only users will see the App in the store but will also install it and use it for a long period of time. (Retention of 15-30 days).

Conclusion

• Reviews: Overall, B4 Bank App users are well satisfied with the products / services offered. We can see a high volume of 5-star ratings as well as the dominance of words that express positive feelings when we analyzed the texts of the reviews.
• Although those who have downloaded the app is satisfied, there is a low conversion of users who visit the store and the ones that actually download the app. To attack this problem, I suggest a deeper analysis of users' behavior as:

Data Science Salary Estimator

To help data scientists better negotiate their income when they get a job.

• Created a tool that estimates data science salaries (MAE ~ $ 11K)
• Scraped over 1000 job descriptions from glassdoor using python and selenium.
• Engineered features from the text of each job description to quantify the value companies put on the knoledge about some topics: python, excel, aws, sql and spark.
• Optimized Linear, Lasso, and Random Forest Regressors using GridsearchCV to reach the best model.
• Built a client facing API using flask

Business Analysis of a Mobility Company

• Explored the data and based on them; *Created a dashboard with the metrics I believe make sense to their executives. *Dashboard views relevant metrics to: CEO, CFO and COO
• Analyzed each metric to propose improvements, campaighs, focus...
• Explored the feeling of driver and riders.

Power BI

In this session, you can fins some data visualization projects.