jda-69

Inspiration

• The opportunity to work with a large amount of data and get a taste of a real-world example of an ML application
• One Nintendo switch for each participante

What it does

• predict the sales for the next 6 months of cerveceria teothiuacan

How we built it

• with previous knowledge about ML and some python libraries that help up speed up the development process and prototyping

Challenges we ran into

• lack of memory (minimum 16Gb ram)
• data visualization problems
• lack processing time

Accomplishments that we're proud of

• we finished one whole prediction

What we learned

• We need more resources for training a large dataset
• Implement strategies to work with large datasets, due to stack overflows

What's next for jda-69

[] model training [] distribute frontpage

Built With

• keras
• numpy
• react
• sklearn