Releases: banishjeffi/mltoolkit
Releases · banishjeffi/mltoolkit
First version of Machie Learning Toolkit
mltoolkit
💫 About Developer:
Banish J
🤖 AI & ML Developer | Data Science & Analytics Expert
🌐 Web Apps & IoT Skilled | Awesome UI Creator
💻 AI is my main focus! 👾
📞 Contact
☎️ 9444333914
📧 mail@banish.in
🌐 Banish
Core Concept
This python package is mde to simplify your ml tasks and make you to run large program at few lines of code it simplifies your program and imprpve your productivity.
Algorithm accuracy table
from mltoolkit import Regression as regressor
acc_table, best_param = regressor.svm(independent, dependent)
acc_table
output
C | linear | rbf | poly | sigmoid | |
---|---|---|---|---|---|
1 | C 10 | 0.022506 | -0.08521 | -0.082239 | -0.099652 |
2 | C 100 | 0.563729 | -0.113243 | -0.084659 | -0.132517 |
3 | C 500 | 0.64177 | -0.10929 | -0.064037 | -0.582106 |
4 | C 1000 | 0.669795 | -0.102105 | -0.032889 | -2.022042 |
5 | C 2000 | 0.767813 | -0.090715 | 0.02603 | -6.818809 |
6 | C 3000 | 0.764471 | -0.079182 | 0.083426 | -14.702022 |
7 | C 7000 | 0.734434 | -0.028374 | 0.291094 | -73.122034 |
how to use
from mltoolkit import Regression as regressor
acc_table, best_param = regressor._algname_(independent, dependent)
acc_table
-
replace
_algname_
withsvm, decision_tree, random_forest, knn
for Regression -
replace
_algname_
withdecision_tree, random_forest, knn
for Classification -
replace
Regression
withclassifier
if its a classification problem statement
Model accuracy table
from mltoolkit import Regression as regressor
reg_report = regressor.fit_model(independent, dependent)
reg_report
output
Metrics | Random Forest | Linear Regression | Poisson Regression | Decision Tree | Support Vector Machine | KNN | |
---|---|---|---|---|---|---|---|
1 | MSE | 20770567.875901 | 32304679.499094 | 30757741.967819 | 45999841.979685 | 172821773.971895 | 112965815.146866 |
2 | MAE 4557.473848 | 5683.720568 | 5545.966279 | 6782.318334 | 13146.169555 | 10628.537771 | |
3 | R2 2714.117549 | 3985.71256 | 3748.157793 | 3099.796517 | 8532.534486 | 7417.95403 | |
4 | RMSE 0.868069 | 0.794807 | 0.804632 | 0.707817 | -0.097732 | 0.282462 | |
5 | R2ADJ | 0.867407 | 0.793777 | 0.803653 | 0.706352 | -0.103238 | 0.278863 |
how to use
from mltoolkit import Regression as regressor
reg_report = regressor.fit_model(independent, dependent)
reg_report
-
replace
Regression
withclassifier
if its a classification problem statement -
replace
fit_model
withfit_save
to save the best model