Application of machine learning algorithms is cumbersome task. The process requires deep understanding of mathematical background, selection, and application of different algorithms. To simplify this, our team came up with the idea of developing the Machine Learning Workbench, as our final year project. Our system aims to provide extensive support for the complete process of the experimental data under consideration including; dataset selection, preprocessing, and application of classification, clustering, and regression’s algorithms. Also, this diverse and comprehensive toolkit can be accessed through a common interface so that users can compare different methods and identify those which are most appropriate for the problem at hand.
OBJECTIVES OF THE SYSTEM:
Our proposed system has the following objectives:
· Provide users the facility to select and visualize the dataset.
· Prepare the dataset by applying different preprocessing techniques.
· Save the preprocessed dataset so it can be used time and again.
· Apply suitable machine learning algorithms and compare their results to know which one best suits the problem at hand.
· Improve the efficiency of result in terms of time and accuracy.
SOCIO-ECONOMIC BENEFITS:
Recently, users have to use one platform to preprocess their dataset and then another to apply machine learning algorithm, that too after understanding the complex mathematics and application standards of each algorithm. To simplify this, we developed and designed the system where all the aforementioned modules can be done through common interface. The system will be free to use and can be used by technical as well as nontechnical users.
PROJECT METHODOLOGY:
Our system aims to provide extensive support for the complete process of the experimental data under consideration including; preparing the input dataset by applying different pre-processing techniques, with an additional privilege to save the pre-processed dataset which could be used time and again. The system also evaluates the learning schemes statistically by plotting them in graph to better understand the graphical critics. Also, it visualizes the output dataset and the result of learning as well as a wide variety of other learning algorithms which includes all the major algorithms of classification, regression and clustering to compare their results so that users know which algorithm best suits their problem.
PROJECT OUTCOME:
Our system will help users to understand, explore and experiment with various machine learning algorithms. Also, the hands on experience that users will get from this system will enable them to hone their skills in this vast and emerging field of machine learning.
