Chen Bocheng

The solution to a kind of Markov chain CRM model with a constant state keeping probability

Based on the solution to a kind of Markov chain (MC) customer relationship management (CRM) model, the resolve solution of the MC CRM model with the state keeping probability is studied in the paper. By computing the inverse matrix for the equation given, we make a resolve solution for this kind of MC CRM model, then shortly make a comparison between the solutions and CLV (the total expected net present value in customer life cycle) results, with and without such state keeping probability.

The expected net present value solution and their CLV comparison of two kinds of Markov CRM models

Pfeifer model and its variation is the Markov chain(MC) models which should be studied in CRM(customer relationship management) analysis. We study the two kinds of model, in infinite horizon purchasing case, demonstrate their expected net present value (ENPV) solution, make a computing to their corresponding CLV (The total expected net present value in customer life cycle), then make a comparison to the CLV of the two scenarios. The study shows: the CLV will vary with the state keeping probability (SKP) changing, and the CLV in the scenario with the SKP will be better than that of without the item.

Optimal production with emission permits and trading under the restricted by government subsidies and punishment mode

Based on the result in literature[5], the paper focuses on exploring enterprise optimal production problem, considering emission permits and trading, restricted by Government subsidies and punishment policy. We build the corresponding model considering such condition, and compute its optimal solution, then investigate the relationship between optimal output and enterprise optimal revenue in the model, also analyze enterprise production optimal management in such restriction. The result shows: the structure of enterprise optimal production and optimal revenue is similar to that in the literature[5], and the optimal revenue is still proportional to the square of optimal production, in the case of linear inverse demand function, only the content in optimal production is a little different.

An improvement on the basic Least Squares Support Vector Machine algorithm

The study on the Least Squares Support Vector Machine (LS-SVM) algorithm is currently a ‘hot’ area of Support Vector Machine (SVM) study. Following Chuas work and the algorithm analysis thought already used in System Identification area, we make a further study to the basic LS-SVM, and make the algorithm more convenient to analyze the estimated parameters, as well as extend his thought to reduce the computing in the enlarged matrix dimension and new data adding cases. A view point is also shown in the paper, there are many similar analysis methods in System Identification theory suit for LS-SVM analysis, and researchers can get twice the result with half the effort if utilize them intentionally when in study.

An improved continuous Hopfield neural network: computing LS parameters and lost function as well as multi-order identification

We focus on an improved structure for the continuous Hopfield neural network (CHNN) in order to give an output of least square (LS) parameters and its corresponding lost function simultaneously. Based on the structure and by controlling the connection status of the corresponding feedback loop of the network on and off, we can compute the LS parameters and lost functions for each LS model order which is lower than the input matrix. Also the improved structure can make the CHNNs fixed structure adapt to any input matrix which may have dimensions lower than or equal to the networks dimensions. Simulation results show that the improved structure can support the functions we have given.