Generative adversarial network (GAN) and enhanced root mean square error (ERMSE): deep learning for stock price movement prediction

The prediction of stock price movement direction is significant in financial circles and academic. Stock price contains complex, incomplete, and fuzzy information which makes it an extremely difficult task to predict its development trend. Predicting and analysing financial data is a nonlinear, time-dependent problem. With rapid development in machine learning and deep learning, this task can be performed more effectively by a purposely designed network. This paper aims to improve prediction accuracy and minimizing forecasting error loss through deep learning architecture by using Generative Adversarial Networks. It was proposed a generic model consisting of Phase-space Reconstruction (PSR) method for reconstructing price series and Generative Adversarial Network (GAN) which is a combination of two neural networks which are Long Short-Term Memory (LSTM) as Generative model and Convolutional Neural Network (CNN) as Discriminative model for adversarial training to forecast the stock market. LSTM will generate new instances based on historical basic indicators information and then CNN will estimate whether the data is predicted by LSTM or is real. It was found that the Generative Adversarial Network (GAN) has performed well on the enhanced root mean square error to LSTM, as it was 4.35% more accurate in predicting the direction and reduced processing time and RMSE by 78 s and 0.029, respectively. This study provides a better result in the accuracy of the stock index. It seems that the proposed system concentrates on minimizing the root mean square error and processing time and improving the direction prediction accuracy, and provides a better result in the accuracy of the stock index.

     

Monitoring surface water content using visible and short-wave infrared SPOT-5 data of wheat plots in irrigated semi-arid regions

Irrigated agriculture is an important strategic sector in arid and semi-arid regions. Given the large spatial coverage of irrigated areas, operational tools based on satellite remote sensing can contribute to their optimal management. The aim of this study was to evaluate the potential of two spectral indices, calculated from SPOT-5 high-resolution visible (HRV) data, to retrieve the surface water content values (from bare soil to completely covered soil) over wheat fields and detect irrigation supplies in an irrigated area. These indices are the normalized difference water index (NDWI) and the moisture stress index (MSI), covering the main growth stages of wheat. These indices were compared to corresponding in situ measurements of soil moisture and vegetation water content in 30 wheat fields in an irrigated area of Morocco, during the 2012–2013 and 2013–2014 cropping seasons. NDWI and MSI were highly correlated with in situ measurements at both the beginning of the growing season (sowing) and at full vegetation cover (grain filling). From sowing to grain filling, the best correlation (R² = 0.86; p < 0.01) was found for the relationship between NDWI values and observed soil moisture values. These results were validated using a k-fold cross-validation methodology; they indicated that NDWI can be used to estimate and map surface water content changes at the main crop growth stages (from sowing to grain filling). NDWI is an operative index for monitoring irrigation, such as detecting irrigation supplies and mitigating wheat water stress at field and regional levels in semi-arid areas.     

Application of Soft Actuation to Bilateral Control and Haptic Reproduction

International Journal of Control, Automation and Systems - The main research topic of this paper is to apply the sliding mode based soft actuation to smooth transition between position, force, and...     

Pulmonary Diffuse Airspace Opacities Diagnosis from Chest X-Ray Images Using Deep Convolutional Neural Networks Fine-Tuned by Whale Optimizer

Wireless Personal Communications - The early diagnosis and the accurate separation of COVID-19 from non-COVID-19 cases based on pulmonary diffuse airspace opacities is one of the challenges facing...     

Porous Silicon： Volume-Specific Surface Area Determination from AFM Measurement Data

Porous silicon layers manufactured by using （100）, 1-5 ohm-cm p-type （boron doped） wafer by electrochemical etching in HF etanol solution. Photoluminescence （PL） spectra of anodically etched silicon obtained for different conditions studied and surface characteristics are investigated by AFM. This study gives a simple way to determine specific surface are of porous silicon which plays a major role with porosity for explaining the blue shift in photoluminescence peak. Properties such as specific surface area, pore size, and pore size distribution, the main surface properties of layer are investigated from AFM data which are important material characteristics in many processing applications. The ＂specific surface area＂ （Sspecific） generally defined as the area of solid surface per unit mass of material, solid volume or cross section area. From 3-D reconstructions of AFM data, the surface area and the volume of the porous layer can be estimated directly and volume-surface specific area is calculated. For porous silicon this feature can be defined as the total surface area per volume and given by the unit m2/cm3. The method is simple not need to construct a special set up for measurement and non destructive.     

Phase Transformation Behavior of Porous TiNi Alloys Produced by Powder Metallurgy Using Magnesium as a Space Holder

Porous TiNi alloys with porosities in the range of 51 to 73 pct were prepared successfully applying a new powder metallurgy fabrication route in which magnesium was used as a space holder, resulting in either single austenite phase or a mixture of austenite and martensite phases dictated by the composition of the starting powders, but entirely free from secondary brittle intermetallics, oxides, nitrides, and carbonitrides. Since transformation temperatures are very sensitive to composition, deformation, and oxidation, for the first time, transformation temperatures of porous TiNi alloys were investigated using chemically homogeneous specimens in as-sintered and aged conditions eliminating secondary phase, contamination, and deformation effects. It was found that the porosity content of the foams has no influence on the phase transformation temperatures both in as-sintered and aged conditions, while deformation, oxidation, and aging treatment are severely influential.     

Self Organized Network Management Functions for Energy Efficient Cellular Urban Infrastructures

Energy efficiency is a significant requirement for the design and management of mobile networks and has recently gained substantial attention from both network operators and the research community. The general concept of energy saving management aims to match the capacity offered by operators to the actual demand at given times and geographic areas. This paper introduces the notion of energy partition, an association of powered-on and powered-off BSs to deliver network-level energy saving. It then elaborates how such concept is applied to perform energy re-configuration to flexibly re-act to load variations encouraging none or minimal extra energy consumption. A simulation-based study evaluates the performance of the proposed algorithms under different network topologies and traffic conditions, highlights the benefits and drawbacks, and provides recommendations for deployment scenarios.     

Proposal of a multi-agent model for the sustainable truck scheduling and containers grouping problem in a Road-Rail physical internet hub

Physical Internet (PI) was introduced as a global standardised and interconnected logistics system based on PI-nodes, PI-movers and PI-containers as a mean toward global logistics sustainability. One important issue regarding PI-nodes concerns the planning and scheduling of operations and the management of PI-containers, both in a deterministic and a perturbed environment. This research considers the Road-Rail PI-hub sustainable truck scheduling and PI-containers grouping problem. In our research we consider the weighted sum of the number of used wagons, the internal distance travelled by PI-containers from PI-docks to wagons as well as the trucks’ tardiness, which translate the search for sustainable logistics. In this paper, an effective and reactive multi-agent system based model (MAS) is developed for the resolution of the trucks scheduling and PI-containers grouping. To ensure the efficiency of the MAS and improve the quality of each of its solutions, three concurrent hybrid meta-heuristics are embedded within three parallel scheduling agents. Then, a mixed integer linear programming model (MILP) is proposed to evaluate the performance of the MAS. Finally, the MAS is also evaluated under internal perturbations. The obtained results show the ability of the MAS to provide alternative sustainable solutions by rescheduling trucks in case of disruptions.