MAADSBML ALGORITHMS

MAADSBML will perform advanced algorithms for both discreet and continous dependents variable.

It will automatically hypertune the parameters of the linear and non-linear algorithms. It will also standardize all the variables.

Important

MAADSBML uses a multi-agent framework to find the optimal algorithm for the data. This is a powerful approach find the optimal algorithm in minutes, when it can normally take days and weeks.

MAADSBML Algorithm Details

MAADSBML will automatically apply the algorithms below to your data. It will know which algorithms to apply to your data based on whether your dependent variable is continuous or discreet.

MAADSBML will also perform cross-validation to hypertune the parameters in the algorithms, where applicable.

Algorithm	Explanation
Linear Regression	This is the standard linear regression algorithm using Ordinary Least Squares (OLS).
Adaboosting	MAADSBML will apply the Adaboosting regressor and Adaboostng classifier to your data.
Neural networks	MAADSBML will apply the neural network algorithm to your data.
SVR/SVC	MAADSBML will apply SVR algorithm to your data and SVC algorithm to your data.
Logistic	For classification models, MAADS will apply several classification algorithm including Logistic Regression.
Gradient boosting	MAADSBML will apply gradient boosting regressor and gradient boosting classifier to your data.
Multiple layer perceptron (MLP)	MAADSBML will apply MLP regressor and MLP classifier to your data.
Ridge regression	MAADSBML will apply the Ridge regression algorithm to your data.
ARIMA	MAADSBML will apply the ARIMA algorithm to your data.
Gaussian	MAADSBML will apply Gaussian process regression and Gaussian process classifiers
Ensemble	MAADSBML will also apply advanced ensemble models.
XGboost	MAADSBML will apply XGBoost algorithm to your data.
Theilsen Regressor	MAADSBML will apply Theilsen regression to your data.
Other special algorithms	Other specialized algorithms are chosen by MAADSBML automatically to ensure users get the BEST algorithm that BEST fits their data.