HWK 3: Classification homework

Hello, dear friend, you can consult us at any time if you have any questions, add WeChat: daixieit

HWK 3: Classification homework - due on 06/14/2024 at 11:59 pm

(The 3rd  part is based on the Long-jump and the Flowers classification example discussed in class):

1) FORECASTABLE currency pair (CP),

2) NON-FORECASTABLE CP and

3) PARTIALLY FORECASTABLE CP.

See data at:

https://drive.google.com/drive/folders/16d0qAeVdxZMC2Sd3yWjgUxTFD2dZeDrS?usp=share_link

1. Organize the data in both an auxiliary MongoDB and MySQL:

a.    Hourly basis

b.    Same starting date

c.    Same ending date

d.    Hourly price (VWAP)

e.    Hourly liquidity (number of transactions)

2.    Every 6 hours (i.e., four times a day), create a vector of features and save it in across-referenced main MongoDB and MySQL (e.g., from auxiliary MongoDB to main MySQL, and vice-versa):

a.    Timestamp

b.   VWAP price

c.    Liquidity (average number of transactions per hour)

d.    (Normalized) Volatility (like in HWK 2)

e.    Max (like in HWK 2)

f.     Min (like in HWK 2)

g.    (Normalized) FD (like in HWK 2): use the first 6-hourslot to generate the Keltner Bands and a distance between 2 bands that make sense (see class explanation).

3.    Use the 70-30 train-test rule to classifyieach currency pair over a regression (use PyCarat) - feel free to use a multi-folder process if necessary.

4.    Write a report (one-page maximum) considering the top-4 Best Practices (Class #3) and explaining the classification and regressions used (please explain). Then, submit it on Brightspace with the codes and the *.csv file you downloaded from the main DBs.