Training a Random Forest

MGMT 638: Data-Driven Investments: Equity

Kerry Back, Rice University

Outline

• Suppose we're happy with our random forest model.
• Then, we should train it using our most current data and save the trained model.
• We can load the trained model and use it to make predictions whenever we want.

Read data

import pandas as pd

# change path_to_file to "./" if the file is in your working directory
path_to_file = "../../"

df = pd.read_csv(path_to_file + "data-2023-11-08.csv")
df.head()

	ticker	date	marketcap	pb	ret	mom	volume	volatility	roe	accruals	agr	sector
0	AACC	2011-01-14	188.3	1.4	-0.014634	-0.184615	2.078000e+04	0.071498	-0.118239	-0.182275	0.004383	Financial Services
1	AAI	2011-01-14	1012.1	2.0	0.002677	0.438224	2.775580e+06	0.128450	0.108073	-0.220045	0.002237	Industrials
2	AAIC	2011-01-14	189.3	1.0	-0.010119	0.684547	3.466000e+04	0.048505	0.136967	0.108055	0.135697	Real Estate
3	AAON	2011-01-14	479.4	4.2	0.007778	0.528685	2.817291e+05	0.044912	0.191801	-0.088557	0.011656	Basic Materials
4	AATC	2011-01-14	63.3	1.4	-0.013960	0.008216	6.800000e+03	0.049756	0.072269	-0.008792	0.089436	Technology

Define model and target variable

from sklearn.ensemble import RandomForestRegressor
forest = RandomForestRegressor(max_depth=3)

df["target"] = df.groupby("date", group_keys=False).ret.apply(
    lambda x: x - x.median()
)

Define predictors (features)

features = [
    "marketcap", 
    "pb", 
    "mom", 
    "volume", 
    "volatility", 
    "roe", 
    "accruals"
]

Filter to most recent 3 years

dates = df.date.unique()
dates.sort()
df = df[df.date.isin(dates[-156:])]

Train the model

forest.fit(X=df[features], y=df.target)

RandomForestRegressor(max_depth=3)

In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook.
On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.

Save the model

from joblib import dump
dump(forest, path_to_file + "forest.joblib")

['../../forest.joblib']