Forecasting demand of short life cycle products by SVM

Xu, Xianhao; Zhang, Hao

doi:10.1109/icmse.2008.4668939

Cited by 4 publications

(5 citation statements)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Kurawarwala et al (1996Kurawarwala et al ( , 1998) considered a seasonal factor on the basis of computer industry [1,5]. Xu et al (2008) analyzed the characteristics of short life cycle products and introduced a bass model [6]. An improved bass model was proposed after feasibility analysis.…”

Section: Literature Reviewmentioning

confidence: 99%

An Improved Forecasting Algorithm for Spare Parts of Short Life Cycle Products Based on EMD-SVM

Fan

et al. 2013

2013 International Conference on Information Science and Cloud Computing Companion

View full text Add to dashboard Cite

Demand of spare parts of short life cycle products has great random fluctuation and short life cycle. Traditional forecasting methods have low forecasting accuracy which leads to understock or overstock of spare parts. Considering such situation an improved forecasting method based on Empirical Mode Decomposition and Support Vector Machine (IEMD-SVM) is proposed. By replacing the Cubic Spline Interpolation in the standard EMD with Piecewise Cubic Hermite Interpolation, the overshoots and undershoots problems caused by great volatility of data are solved. Experiments with 459 real data sets show that the IEMD-SVM forecasting method has a better forecasting result than traditional forecasting methods which provides better decision supports for enterprise inventory management.

show abstract

Section: Literature Reviewmentioning

confidence: 99%

An Improved Forecasting Algorithm for Spare Parts of Short Life Cycle Products Based on EMD-SVM

Fan

et al. 2013

2013 International Conference on Information Science and Cloud Computing Companion

View full text Add to dashboard Cite

show abstract

“…SVM is especially useful for nonlinear, high dimension and local minimum problem. According to the references [3,5,7,8], the principle of SVR is as follows.…”

Section: Support Vector Regressionmentioning

confidence: 99%

“…Suppose x i ∈R are main factors that affect cashmere price, and y i is the price forecasting value. Cashmere price can be expressed by the features, which is given in (7) ( , , ) The goal for SVM based price forecasting model is to find the relationship between i x and i y .So according to the theory presented in section 2, price forecasting model is establish as follows:…”

Section: Establish Cashmere Price Forecasting Modelmentioning

confidence: 99%

Parameter Selection of SVR Based on Improved K-Fold Cross Validation

Wang

Qiao

2013

AMM

View full text Add to dashboard Cite

This article firstly uses svm to forecast cashmere price time series. The forecasting result mainly depends on parameter selection. The normal parameter selection is based on k-fold cross validation. The k-fold cross validation is suitable for classification. In this essay, k-fold cross validation is improved to ensure that only the older data can be used to forecast latter data to improve prediction accuracy. This essay trains the cashmere price time series data to build mathematical model based on SVM. The selection of the model parameters are based on improved cross validation. The price of Cashmere can be forecasted by the model. The simulation results show that support vector machine has higher fitting precision in the situation of small samples. It is feasible to forecast cashmere price based on SVM.

show abstract

“…Traditional forecasting methods are often inappropriate for estimating demand for a product with a short life cycle because they do not take into account the characteristics of the product's life cycle and usually require a significant demand history, which is available only after the product has been sold for some time (Zhu and Thonemann, 2004). For short life cycle products, little or no historical data is available for forecasting demand, which makes it very difficult to predict demand (Kamath and Roy, 2007;Xu and Zhang, 2008). A useful demand prediction system must accommodate the unique characteristics of the short life cycle product.…”

Section: Characteristic Of Products (Short Life Cycle Products) and P...mentioning

confidence: 99%

“…Due to the continuous evolution of science and technology, personalized products, and combinations of functional products with innovative products, products upgrade with increasing frequency. New production technology is introduced often and more products have the characteristics of short life cycles, especially high tech goods, consumer electronics, and personal computers(Kurawarwala and Matsuo, 1996;Xu and Zhang, 2008;Aytac and Wu, 2011). Both capacity and investment in capacity are costly, and firms must avoid over-investing in capacity,…”

mentioning

confidence: 99%

Long term capacity planning with products' renewal

Yilmaz

View full text Add to dashboard Cite

Long Term Capacity Planning (LTCP) consists of deciding the type and amount of capacity of production systems for multiple periods in a long term planning horizon. It involves decisions related to strategic planning, such as buying or selling of production technology, outsourcing, and making tactical decisions regarding capacity level and configuration. Making these kinds of decisions correctly is highly important for three reasons. Firstly, they usually involve a high investment; secondly, once a decision like this is taken, it cannot be changed easily (i.e. they are highly irreversible); thirdly, they affect the performance of the entire system and the decisions that will be possible at a tactical level. If capacity is suboptimal, there will be lost demand (in the present and possibly in the future); if the system is oversized, there will be unused resources, which may represent an economical loss. Long term decisions are typically solved with non-formalized procedures, such as generating and comparing solutions, which do not guarantee an optimal solution. In addition, the characteristics of the long term capacity planning problem make the problem very difficult to solve, especially in cases in which products have a short life cycle. One of the most relevant characteristics is the uncertainty inherent to strategic problems. In this case, uncertainty affects parameters such as demand, product life cycle, available production technology and the economic parameters involved (e.g. prices, costs, bank interests, etc.). Selection of production technology depends on the products being offered by the company, along with factors such as costs and productivity. When a product is renewed, the production technology may not be capable of producing it; or, if it can, the productivity and/or the quality may be poor. Furthermore, renewing a product will affect its demand (cannibalization), as well as the demand and value of the old products. Hence, it is very important to accurately decide the correct time for product renewal. This thesis aims to design a model for solving a long term capacity planning problem with the following main characteristics: (1) short-life cycle products and their renewal, with demand interactions (complementary and competitive products) considered; (2) different capacity options (such as acquisition, renewal, updating, outsourcing and reducing); and (3) tactical decisions (including integration strategic and tactical decisions).

show abstract

Forecasting demand of short life cycle products by SVM

Cited by 4 publications

References 13 publications

An Improved Forecasting Algorithm for Spare Parts of Short Life Cycle Products Based on EMD-SVM

An Improved Forecasting Algorithm for Spare Parts of Short Life Cycle Products Based on EMD-SVM

Parameter Selection of SVR Based on Improved K-Fold Cross Validation

Long term capacity planning with products' renewal

Contact Info

Product

Resources

About