Empirical Study on Collaborative Software in the field of Machine learning.pptx

Empirical Study on Collaborative Software in the field of Machine learning

1. Empirical Study on Collaborative Software in the field of Machine learning based Approach on Image Processing By Waqar Khan Class Name: 协同软件技术及应 By Professor: 申利民

2. Agenda Collaborative software Model Framework Problem Statement Introduction Methodology Conclusion

3. Collaborative Software

4. Data Training Team Deployment

5. Model Framework

6. Problem Statement Classification between the objects is easy task for humans but it has proved to be a complex problem for machine. The raise of high capacity computers, the availability of high quality and low priced images. The increasing need for automatic image analysis has generated and interest in images classification algorithms for better software.

7. Introduction System consists of database that contains predefined patterns that compares with detected object to classify in the proper category.

8. Methodology Feature Mining :- The concept of feature mining, in which our goal is to minimize the human effort needed to explore and organize the vast space of possible features for image classification. Parameterized Feature Space P Data Driven Feature Space F X represents the data S pace Parameterized feature space P f is a function of f: X          Where error of training data,  

9. Ada boost classifier Given N labelled training examples (x i , y i ) with y i ϵ {-1,1}, and x i ∈X , and an initial distribution D i ( i ) over the examples, AdaBoost combines a number of weak classifiers h t to learn a robust classifier H(x)=sign(f(x)). The training error    

10. Model Evaluation Here are some expected experimental results on datasets. The P which represent the feature space, little efforts for P because it must be large and complex enough to represent diverse patterns. Even for a 50x50 image patch, there are ϑ(10^6 ) configurations for a single rectangle. With multiple rectangles per Haar wavelets, P becomes quite vast.

11. Conclusion Several teams have devoted significant effort to establishing a comprehensive portfolio of AI research based on image classification and other areas of images. Feature mining alleviates the effort and expertise necessary for feature design. A general framework for feature mining in images, grounding it in theory and supporting expected results.

12. Thank you