Information Extraction And Parametric Analysis Of LED

Data mining techniques emerge by the arrival of the information age.With the developmentof computer techniques and data warehouse, they are almost applied to various fields, such assales, telecommunications, financial services, insurance and many other fields.Semiconductorlight-emitting diode is a new green light source.It maintains a strong market competitiveness byits unique advantages.This paper analyzes the correlated color temperature of LED, fits and clusters the correlatedcolor temperature’s curve which changed with the aging time.It Provides a strong basis forreliability assessment and screening of LED. Fitting the curve of LED’s correlated colortemperature by approximation methods,such as Fourier, Gaussian, Multi-form and Sine curvefunctions,and analysing the results.Clustering the correlated color temperature by hierarchical,K-means, fast clustering and FCM. Testing and verifiing K-means by the software of SPSS.Developping the Human-Computer Interaction by Visual Studio in order to achieve the purposeof importing date, fitting and clustering.