SystemsLogical

SystemsLogical / ClassificationScorer / 0.1.7

README.md

Overview

This algorithm is NOT ready for prime-time, launch date: 4/1/2018

Runs input through five scikit-learn classification models, scores, and reports best results for data input

Applicable Scenarios and Problems

Fast and simple way to choose the correct classification model for your particular data set. Why waist your valuable time coding, instatiating, fitting, predicting, scoring, and evaluating multiple models just to find throw that work away in favor of the single best. For just a few pennies, find the best model for your dataset.

Input

Two .csv files are required for input. The FIRST must be the features The SECOND must be the targets Both files must not have a header row

These input files must be fully feature engineered, clean files ready for fitting and predicting by Scikit-Learn classification models.

Currently these five models are instatiated, fitted, predicted, and scored:

**Decision Tree

**Support Vector Machine (SVM)

**K Nearest Neighbors (KNN)

**Perceptron

**GaussianNB

Examples (python)

Example 1:

client = Algorithmia.client('your_API_key_here');

feature_file = "data://SystemsLogical/ml_data/diabetes_features.csv";

target_file = "data://SystemsLogical/ml_data/diabetes_targets.csv";

input1 = [feature_file, target_file]

algo = client.algo('SystemsLogical/ClassificationScorer/0.1.7')

result = algo.pipe(input1)

print(result.result)

a log file called classifier-output.txt (the output of the algorithm) has been placed in a temp algorithm directory in your account

Example 2:

client = Algorithmia.client('your_API_key_here');

input2 = ["data://FifeMeister/ml_data/diabetes_features.csv", "data://FifeMeister/ml_data/diabetes_targets.csv"]

algo = client.algo('SystemsLogical/ClassificationScorer/0.1.7')

result = algo.pipe(input2)

print(result.result)

a log file called classifier-output.txt (the output of the algorithm) has been placed in a temp algorithm directory in your account

Output

The output is a log file called classifier-output.txt which is placed in the caller's data collection, within an algo/temp directory.

data://.algo/your_user_name/ClassificationScorer/temp/classifier-output.txt