/Data-Science-Project

Forecasting restaurant customer demand with LSTM, 1DConvNet, and DNN.

Primary LanguagePython

Data Science Project for Cross Cafe - Aarhus

Objective:

Inducing a more data-driven decision making in a café environment.

Cafés are typically small businesses, most café owners, own only a few cafés with little knowledge and resources for inducing data-driven decisions. This project would help a particular café extract it’s data from various systems and collect it in a dashboard to easily investigate when making key changes in the business. The main goal of which is to help the café get a better overview of their operation while also providing future insights through forecasting that help them make critical decisions looking forward.

Data sources:

  • Data from the café

  • Potential interviews or other qualitative input

  • Review data from various websites

Wind data observations:

Available (almost) every hour:

Cloud_cover - (0-100%)

Humidity - (0-100%)

Precip_dur_past1h - (0-60 min)

Precip_past1h - (precip in mm)

Pressure - (some missing data / different methods of measurements)

Temp_dew - (degrees c - suppose it is a measure of humidity, but not sure)

Temp_dry - (degrees c - “air temperature”- some way of disregarding humidity’s effect on the temperature(?))

Temp_max_past1h - (degrees c)

Temp_mean_past1h - (degrees c)

Temp_min_past1h - (degrees c)

Wind_dir - (in degrees (0-360) - maybe transform to categorical variable)

Wind_max_per10min_past1h - (m/s)

Wind_speed - (m/s)

Wind_speed_past1h - (m/s)


Available more than once a day, but not hourly:

Weather (no idea to be honest, values go from 100-185)

Temp_max_past_12h (degrees c, measured at 6:00 and 18:00)

Temp_min_past 12h (degrees c, measured at 6:00 and 18:00)

Wind_min_past1h (m/s)


DMI weather descriptions

https://confluence.govcloud.dk/pages/viewpage.action?pageId=26476616&fbclid=IwAR3IIvZ6_gABMYjlUJ4lwSZX545XY0DEKGIBxVj13utg8Wmuky5w-9DUyS8

written by:

Timon Florian Godt, Morten Hamburger, Daniel Bolander, Piratheban Rajasekaran