BWIQ®

Graphical, Logical Chemometrics for Spectroscopic Data BWIQ® is a comprehensive multivariate analysis software package for analysis of spectral data including exploratory, qualitative analysis and quantitative regression methods.

BWIQ supports the classical chemometric methods for classification and regression including Partial Least Squares Regression (PLS), Principal Component Analysis (PCA) Software

and Discriminant Analysis with Support Vector Machine (SVM) algorithms for non-linear datasets. It features an intuitive workflow from data import to prediction, with extensive graphics in support of model analysis and interpretation. BWIQ includes preprocessing tools for optimal analysis of Raman, NIR, LIBS and other spectroscopic datasets. The BWIQ chemometrics software package is ideal for online use with the i-Raman® series instruments for real-time prediction and offline use with high-resolution spectroscopic data.

Applications:

• Exploratory Data Analysis

• Multivariate Quantitative Analysis

• Multivariate Classification Analysis

• Real-time Prediction of New Data Key Features:

• Full suite of regression and classification routines

• Extensive graphical display of data and results for evaluation of model performance metrics

• Easy import of BWSpec®, spc, Matlab and csv data file formats

• Logical easy-to-follow work flow and progressive software structure

• Chemometric Modeling Markup Language (CMML) for easy model storage, sharing and use within operating software including BWAnalyst

• Innovative algorithms for baseline correction (airPLS) and spectral smoothing (Whittaker Penalized Least Squared)

• Real-time prediction with i-Raman series instruments

• Report function including model parameter summary

Contact us for a free trial version of BWIQ®





Main Functions:

• Exploratory data analysis through Principal Component Analysis (PCA)

• Regression analysis with various algorithms including MLR, PCR, PLS and Support Vector Machines (SVMR)

• Classification and discriminant analysis with algorithms including SIMCA, PCA-MD, PLS-DA, and Support Vector Machine Classification (SVC)

• Sample partition algorithms for sample selection including Kennard-Stone and SPXY

• Numerous spectral preprocessing algorithms, such as automatic baseline correction, smoothing, derivatives, and normalization

• Cross-validation and test set validation options

• Outlier detection using Y-residuals, Q-residuals, M-distance

• Full model details table viewable within model

• Portability of cmml model file use with BWIQ prediction engine in BWAnalyst software System requirements:

Example applications:

The following hardware and software are

• Quantitative analysis of API in pharmaceutical tablets recommended for optimal performance of BWIQ:

• Product contamination: quantitation of diethylene Operating Systems

glycol in glycerin; methanol contamination in alcoholic

• Windows 7, 8, or 10

beverages

Recommended Hardware

• Quantitation of fatty acids in edible oils

• Intel Core i7 or greater

• Real-time monitoring of glucose in nutrients;

• 4 GB RAM or more

quantitation of glucose in aqueous solutions

• 1 GB hard drive space

• Sample classification: gasoline of different octane levels; alcohols; different edible oils

• monitor with 1024 x 768 pixels or higher

• video card with high GDI+ drawing speed Example Software Work Flow

• Sample selection

Predict unknown samples from stored

• Spectral

data, or in real-time

preprocessing

with SG first

derivative

PLS Regression

Model

19 Shea Way, Newark, DE 19713, USA • Tel: +1 (302) 368-7824 • Web: www.bwtek.com Copyright 2018 B&W Tek LLC · Doc-Rev: 280001218-C (2018/11/29)