A Study On The Human Resources Capital Investment In The Internationalising Enterprises Based On BP Neural Network

By reviewing the developing history of human capital investment of enterprises and concluding the achievements in this field both at home and abroad, this thesis analysed the difficulty of training demand identification and evaluation and built a decision-making system of training demand evaluation in the internationalising corporations on the base of BP neural network, one category of the artificial intelligence, by taking M Corporation as an example. The system is designed on the indices selected by the department of human resources management in M Corporation and the database containing the detailed information of employees in the company, which is constructed by the department of human resources management too. This thesis accomplished the starting steps of a computer-aided analysing frame that was suited for M Corporation, and the other departments in different corporations are able to adjust the default settings according to the available internal data-exchange platform and database system, and then build up and improve the new automatic decision-making systems that are able to be applied to the other fields of business management, so that the corporations hold the ability to identify and seize the flashing chances in the external market and react to the changes swiftly, relying on the huge storage and ultra-high-speed computation of data in computers, which are helpful to enhance the efficiency of making decisions and create more values for the corporations.This thesis contains five chapters, which can be generalised into three parts:Part I is the first chapter. It illustrated the research background of the problems discussed in this thesis, summarised the current situation of cross-cultural management in the internationalising enterprises and stated the innovation of this thesis.Part II is the second chapter, which is the theoretical basis of the research in this thesis. It reviewed the domestic and abroad research achievements of human capital investment in enterprises at the beginning, followed by the cross-cultural management and revealed the deficiencies of the present managing practices in the domestic enterprises. This part also introduced the basic conceptions, generating process, frameworks of theories and model construction, the advantages and disadvantages in practice of BP neural network, which is one subset of the artificial intelligence systems.Part III contains the third and fourth chapters. This part started with the influential factors of internationalisation and the human capital investment in internationalising enterprises, which are the basis of the indices system in this thesis, followed by the analysis on feasibility of fuzzy evaluation system by applying BP neural network, including the analysis on internal relationship between the BP neural network and the issue explored in this thesis. In the following, the indices system was completed by clarifying the principles of construction, the components and the requirements to the original data, in addition to the characteristics of structure and elements according to the restrictions of BP neural network, which was followed by the construction, training and simulation on BP neural network computing through computer programmes on the platform of MATLAB, in accordance with the defined indices system and model parameters before comparing the predicted results by the network and the real ones made by experts of the M Corporation. For the sake of the further verification on the correctness of network predictions, the multinomial logistic regression was selected before the terminal of this part to perform a test based on the prediction output of trained BP neural network to a new set of case data, which revealed that the prediction of the network was valid and reliable.Part IV is the fifth chapter. This part concluded the basic idea and main method while constructing the cross-cultural training demand evaluation model for the internationalising enterprises on the base of BP neural network, as well as the possible and plausible directions, fields, methods for extension and the significance to the managing practice of enterprises. At the end of this part, some difficulties and suggestions in application of BP neural network in managing practice were expressed to help the corporations upgrade and optimise the network, so that it could be more suitable for the actual demand in business management.