Makoondlall, Akshay (2019) A pattern recognition system for the automation of learning organisation – Learning Organisation Information System (LOIS). (MSc(R) thesis), Kingston University, .
Abstract
This thesis is about finding a solution to achieve the automation of the Learning Organisation Information System code, named LOIS, where large companies will be able to instigate a culture of team learning based on the analysis of the events that occur in their respective businesses. They will be able to categorise data and formulate measures that can be implemented to reinforce and adapt their business models to accommodate changes. Nevertheless, this can only work if the automation offers non-intrusiveness, ease of use and adaptability for the big companies. The main idea behind implementing LOIS is to provide a platform for employees to express their concerns and predict the number of leavers whereby Organization Management can formulate measures to retain those employees. The solution involves two algorithms; firstly K-means algorithm that is used to cluster data and secondly Time Series Prediction that is used in conjunction to make prediction on those clustered data. It explains how K-means algorithm runs through all the data until a ‘no points change’ cluster membership is reached and how Time Series Prediction is used to clustered data to predict by initially normalizing the set of data and then by fluctuating the number nodes (layers). The main idea of implementing Time Series Prediction into the system is to predict the number of employees that will potentially leave the organization over a certain period. An architectural framework has been incorporated within this thesis that has then been built based on a Case study that has been designed specifically to implement the framework whereby results are generated to be analysed, reviewed and formulate measures. The thesis explains in detail all the different components of the framework, the Process Flow, the Deployment Architecture while concluding with an Organisational Framework Process Flow. Furthermore, it explains and shows the different graphical user interface that the system has to offer to employees in order to help their day to day life within a company. The thesis concludes by comparing a huge amount of historical and actual data into the system to know if there are any improvement in the processes.
Actions (Repository Editors)
 |
Item Control Page |