LearnIt makes the machine learning pipeline easy for you. What you need is just a CSV file and you can quickly obtain a trained machine learning model with evaluation results. You don't have to manually extract features or select machine learning algorithms or tune hyper-parameters or evaluate trained models. LearnIt takes care of everything from end to end.
LearnIt is built on top of Pandas and Scikit-learn, popular data science tools, to minimize additional learning cost. You can also register your own feature extraction templates into the library to make LearnIt better.
You can install LearnIt (learnit) with pip command.
$ pip install learnit
or
$ pip install git+http://git@github.com/megagonlabs/learnit.git
import pandas as pd
from learnit.autoconverter.autoconverter import AutoConverter
from learnit.autolearn.autolearn import AutoLearn
# Load data as pd.DataFrame
df = pd.read_csv("data/train.csv")
# Create AutoConverter instance and convert
ac = AutoConverter(target="Survived")
X, y = ac.fit_transform(df)
# Model selection
al = AutoLearn()
al.learn(X, y)
Please also see a detailed step-by-step tutorial.
- Hideki Awashima (@haisland0909)
- Kate Pavlovic (@katepavlovic)
- Yoshihiko Suhara (@suhara)
Please create GitHub issues and PRs. For private communication, please contact Megagon Labs LearnIt Team (learnit@megagon.ai)