Md. Janibul Alam Soeb1, Muhammad Rashed Al Mamun2
Department of Farm Power and Machinery, Sylhet Agricultural University, Sylhet, Bangladesh.
Abstract: Now a day’s load forecasting leads an immense area of research in power system. It helps to generate the power with minimal cost and ensure the reliability of power systems. It is so much attractive because accurate load forecasting is a challenging task for its difficulties. This paper presents the load forecasting for the Power Grid Company Bangladesh Ltd. (PGCB) by using Advanced Back Propagation Algorithm. It is advanced because here the adaptation mechanism is used to update hidden layer. And it uses newly designed data set where only that kind of inputs are chosen which give the best prediction output. These inputs are chosen by trial and error method. The data are collected from PGCB to train the system. This paper has proposed to train the network in summer for reducing load shedding and in winter, holidays to minimize the power loss as well as the cost of generation. Experimental results show that the system provides the load forecasting with high accuracy.
Keywords: Load forecasting, PGCB, Artificial neural network, Back propagation algorithm.
Pages: 95 – 102 | Full PDF Paper
Emanuele La Malfa1, Gabriele La Malfa2
1. University of Oxford, Oxford, UK.
2. EMLYON Business School, Paris, FRANCE
Abstract: We propose an unsupervised machine learning algorithm for anomaly detection that exploits self-learnt features of monodimensional time series. A Variational Autoencoder, where convolution takes place of dot product, is trained to compress each input to a low-dimensional point from a normal distribution, detecting an anomaly as low probability and high density sequence. We validate our work on different public datasets, obtaining results that shed new light on Variational Autoencoders applied to anomaly detection.
Pages: 103 – 108 | Full PDF Paper
ClinBAY ltd, Cyprus
Abstract: Carryover effects constitute a potential issue when using the crossover design. Multiple methods exist, both for modelling the carryover effect, and predicting adjusted means from these models. In the current paper, we investigated the reliability of the prediction methods for estimating the model adjusted means and their differences, using a simulation study. Our simulation data suggest that the most reliable estimates were obtained when modelling the carryover effect as a factor, and assuming a carryover of zero in the predictions. However, potential confounding effects can cause large bias when using this method, and in the author’s opinion it would be preferable to completely avoid presenting the model adjusted means. Most of the methods tested, as expected, provided identical and reliable estimates of the treatment differences from these model adjusted means, when a carryover was included in the model either as factor, or using a reduced carryover model.
Keywords: Carryover effect, prediction, model adjusted means, LSMEANS, least square means, marginal means.
Pages: 109 – 122 | Full PDF Paper