Predicting inflation in Mexico

• 

  Models prediction

Inspiration

The inspiration for us was the opportunity to apply Nowcasting (as a new topic for us), in the prediction of such an important economic indicator for any country: inflation.

What it does

The built statistical models predict, using data from 2000 to 2020 in a bi-weekly period, the inflation (INPC value) for Mexico in 2021.

How we built it

• Using Python Notebooks, with packages such as Pandas, statsmodels, sklearn, Numpy, Facebook Prophet, and Matplotlib.

• Applying our code to data from INEGI, from this link: https://www.inegi.org.mx/programas/inpc/2018/#Tabulados

Challenges we ran into

The problem comprehension (because it's from a different area than us), and how to determine the best approach for increasing accuracy in our models.

Accomplishments that we're proud of

• The comprehension of an economic problem
• The implementation of Data Science for a real problem in our country.
• Learn and apply different frameworks in such a small window of time.

What we learned

• About how inflation is calculated for a country.
• Nowcasting.
• Time series modeling.

What's next for Predicting inflation in Mexico

• Implement Machine Learning models.
• Try other approaches to solve the problem and increase the accuracy.

Built With

• fbprophet
• matplotlib
• numpy
• pandas
• python
• sklearn
• statsmodels