Welcome to the website of Multimodeler software.
Multimodeler is a data analysis and modeling software with graphical user interface. It is developed considering the need for easy and fast time-series modeling and analysis. Multimodeler provides modules for conventional linear-nonlinear modeling and also neural network-deep learning modeling.
Multimodeler provides the following functions:
1. Data/analysis types: Data can be undated, time-series or panel data.
2. Analysis functions: Multimodeler has the following analysis functions: Descriptive statistics, histogram plots, covariance analysis, autocorrelation, cross-correlation analysis and plots, frequency analysis, augmented Dickey-Fuller (ADF) test, Kwiatkowski-Phillips-Schmidt-Shin (KPSS) test, Phillips-Perron (PP) test, sensitivity analyses.
3. Operations: Mathematical operations on series, seasonal adjustment (additive and multiplicative), detrending, Baxter-King filter, Christiano-Fitzgerald filter, Hodrick-Prescott filter.
4. Statistical and time-series modeling: Linear regression, AR-MA-ARMA-ARIMA-SARIMA-ARIMAX-SARIMAX modeling using OLS solver with manual and auto degree selection, AR-MA-ARMA-ARIMA-SARIMA-ARIMAX-SARIMAX modeling using MLE solver with manual and auto degree selection (BFGS and ABC type optimizers), simple exponential smoothing, Holt’s linear and Holt-Winters modeling, ARCH-GARCH models, VAR-VARMA-VARMAX modeling, Engle-Granger test and ECM modeling, Johansen cointegration test and VECM modeling, state-space modeling with Kalman filters, nonlinear modeling: Cobb-Douglas, exponential growth/decay, logistic (s-curve), constant elasticity of substitution, power model, rectangular hyperbola, shifted hyperbola, Michaelis-Menten modeling.
5. Multilayer Perceptron Type Artificial Neural Network (ANN) modeling: Autoregressive artificial neural network (AR-ANN) model, input-output type artificial neural network (X-ANN) modeling, autoregressive artificial neural network with exogenous variables (ARX-ANN) modeling. These models can have manual or automatic setting for the determination of lag order, neuron number, etc. All ANN models are transparent and can be verified by 3rd party software using scripts exported by Multimodeler easily.
6. Multilayer Perceptron Type Deep Learning Network Network (DLN) modeling: Autoregressive deep learning network (AR-DLN) model, input-output type deep learning network (X-DLN) modeling, autoregressive deep learning network with exogenous variables (ARX-DLN) modeling. These models can have manual or automatic setting for the determination of lag order, neuron number, etc. All DLN models are transparent and can be verified by 3rd party software using scripts exported by Multimodeler easily.
7. Forecasting functionality: All of the modeling panels of the Multimodeler have easy and fast forecasting features.