Search

In time-series forecasting, statistical methods and various newly emerged models, such as artificial neural network (ANN) and grey model (GM), are often used. No matter which forecasting method one would apply, it is always a huge challenge to make a sound forecasting decision under the condition of having very few historical data. Unfortunately, in fashion color trend forecasting, the availability of data is always very limited owing to the short selling season and life of products. This motivates us to examine different forecasting models for their performances in predicting color trend of fashionable product under the condition of having very few data. By employing real sales data from a fashion company, we examine various forecasting models, namely ANN, GM, Markov regime switching, and GM+ANN hybrid models, in the domain of color trend forecasting with a limited amount of historical data. Comparisons are made among these models. Insights on the appropriate choice of forecasting models are generated. [ABSTRACT FROM PUBLISHER]

The Artificial Neural Network (ANN) and its variations have been well-studied for their applications in the prediction of industrial control and loading problems. Despite showing satisfactory performance in terms of accuracy, the ANN models are notorious for being slow compared to, e.g., the traditional statistical models. This substantially hinders ANN model's real-world applications in control and loading prediction problems. Recently a novel learning approach of ANN called Extreme Learning Machine (ELM) has emerged and it is proven to be very fast compared with the traditional ANN. In this paper, an Intelligent Quick Prediction Algorithm (IQPA), which employs an extended ELM (ELME) in producing fast, stable, and accurate prediction results for control and loading problems, is devised. This algorithm is versatile in which it can be used for short, medium to long-term predictions with both time series and non-time series data. Publicly available power plant operations and aircraft control data are employed for conducting analysis with this proposed novel model. Experimental results show that IQPA is effective and efficient, and can finish the prediction task with accurate results within a prespecified time limit. [ABSTRACT FROM PUBLISHER]

Abstract: Forecasting future color trend is a crucially important and challenging task in the fashion industry including design, production and sales. In particular, the trend of fashion color is highly volatile. Without advanced methods, it is very hard to make fashion color trend forecasting with reasonably high accuracy, and it is a handicap for development of the intelligent expert systems in fashion industry. As a result, many prior works have employed traditional regression models like ARIMA or intelligent models such as artificial neural network (ANN) and grey model (GM) for conducting color trend forecasting. However, the reported accuracies of these forecasting methods vary a lot, and there are controversies in the literature on these models’ performances. As a result, in this paper, we systematically compare the performances of ARIMA, ANN and GM models and their extended family methods. With real data analysis, our results show that the ANN family models, especially for Extreme Learning Machine (ELM) with Grey Relational Analysis (GRA), outperform the other models for forecasting fashion color trend. [Copyright &y& Elsevier]

Technology-mediated collaboration process has been extensively studied for over a decade. Most applications with collaboration concepts reported in the literature focus on enhancing efficiency and effectiveness of the decision-making processes in objective and well-structured workflows. However, relatively few previous studies have investigated the applications of collaboration schemes to problems with subjective and unstructured nature. In this paper, we explore a new intelligent collaboration scheme for fashion design which, by nature, relies heavily on human judgment and creativity. Techniques such as multicriteria decision making, fuzzy logic, and artificial neural network (ANN) models are employed. Industrial data sets are used for the analysis. Our experimental results suggest that the proposed scheme exhibits significant improvement over the traditional method in terms of the time–cost effectiveness, and a company interview with design professionals has confirmed its effectiveness and significance. [ABSTRACT FROM PUBLISHER]

Fabric selection is a crucial step in fashion product development. Prior research works have studied the prediction of fabric specimens based on the fabric hand descriptors via either traditional statistical methods or artificial intelligence methods. Despite showing good prediction accuracy, these methods usually lack an understandable ruleset, which means their “interpretability” is low. In this paper, a fuzzy neural network (FNN) based intelligent fabric hand prediction system is explored. Unlike some traditional FNN models in which a full ruleset of the artificial neural network (ANN) is presumed, the proposed FNN system includes a simplification of the network structure and feature selection, so that the number of rules is significantly reduced without big sacrifice on prediction accuracy. Real datasets collected from 30 participants’ evaluation on a set of ten fabric specimens are used to train and test the performance of the proposed system. The system's prediction accuracy is found to be over 80%. Applications of the proposed system are discussed and future research directions are outlined. [ABSTRACT FROM PUBLISHER]

Abstract: Sales forecasting is crucial in fashion business because of all the uncertainty associated with demand and supply. Many models for forecasting fashion products are proposed in the literature over the past few decades. With the emergence of artificial intelligence models, artificial neural networks (ANN) are widely used in forecasting. ANN models have been revealed to be more efficient and effective than many traditional statistical forecasting models. Despite the reported advantages, it is relatively more time-consuming for ANN to perform forecasting. In the fashion industry, sales forecasting is challenging because there are so many product varieties (i.e., SKUs) and prompt forecasting result is needed. As a result, the existing ANN models would become inadequate. In this paper, a new model which employs both the extreme learning machine (ELM) and the traditional statistical methods is proposed. Experiments with real data sets are conducted. A comparison with other traditional methods has shown that this ELM fast forecasting (ELM-FF) model is quick and effective. [Copyright &y& Elsevier]