scholarly journals Predicting the Future is like Completing a Painting: Towards a Novel Method for Time-series Forecasting

IEEE Access ◽  
2021 ◽  
pp. 1-1
Author(s):  
N. Maaroufi ◽  
M. Najib ◽  
M. Bakhouya
Keyword(s):  
Time Series ◽  
Time Series Forecasting ◽  
The Future ◽  
Novel Method

scholarly journals A data-driven method to detect the flattening of the COVID-19 pandemic curve and estimating its ending life-cycle using only the time-series of new cases per day.

2020 ◽  
Author(s):  
Dr. Praveen Gupta ◽  
Prof. K.K. Sharma ◽  
Prof. S.D. Joshi ◽  
Dr. S. Goyal
Keyword(s):  
Time Series ◽  
Life Cycle ◽  
Real Life ◽  
The United States ◽  
Health Care Facilities ◽  
The Future ◽  
Containment Measures ◽  
The Usa ◽  
Novel Method ◽  
Novel Coronavirus

The novel Coronavirus-19 disease (COVID-19) has emerged as a pandemic and has presented itself as an unprecedented challenge to the majority of countries worldwide. The containment measures for this disease such as the requirement of health care facilities greatly rely on estimating the future dynamics and flattening of the COVID-19 curve. However, it is always challenging to estimate the future trends and flattening of the COVID-19 curve due to the involvement of many real-life variables. Recently, traditional methods based on SIR and SEIR have been presented for predictive monitoring and detection of flattening of the COVID-19 curve. In this paper, a novel method for detection of flattening of the COVID-19 curve and its ending life-cycle using only the time-series of new cases per day is presented. Simulation results are compared to the SIR based methods in three different scenarios using COVID-19 curves for South Korea, the United States of America, and India. In this study, simulations, performed on the 26th April 2020 show that the peak of the COVID-19 curve in the USA has already arrived and situated on the 14th of April 2020, while the peak of the COVID-19 curve for India has yet to arrive.

scholarly journals A Decision-Support System for the Digitization of Circular Supply Chains

2021 ◽  
pp. 97-107
Author(s):  
Dimitris Ntalaperas ◽  
Iosif Angelidis ◽  
Giorgos Vafeiadis ◽  
Danai Vergeti
Keyword(s):  
Machine Learning ◽  
Time Series ◽  
Decision Support ◽  
Decision Support System ◽  
Support System ◽  
Learning Model ◽  
Time Series Forecasting ◽  
Machine Learning Model ◽  
The Future ◽  
Fast Processing

AbstractAs it has been already explained, it is very important for circular economies to minimize the wasted resources, as well as maximize the utilization value of the existing ones. To that end, experts can evaluate the materials and give an accurate estimation for both aspects. In that case, one might wonder, why is a decision support system employing machine learning necessary? While a fully automated machine learning model rarely surpasses a human’s ability in such tasks, there are several advantages in employing one. For starters, human experts will be more expensive to employ, rather than use an algorithm. One could claim that research towards developing an efficient and fully automated decision support system would end up costing more than employing actual human experts. In this instance, it is paramount to think long-term. Investing in this kind of research will create systems which are reusable, extensible, and scalable. This aspect alone more than remedies the initial costs. It is also important to observe that, if the number of wastes to be processed is more than the human experts can process in a timely fashion, they will not be able to provide their services, even if employment costs were not a concern. On the contrary, a machine learning model is perfectly capable of scaling to humongous amounts of data, conducting fast data processing and decision making. For power plants with particularly fast processing needs, an automated decision support system is an important asset. Moreover, a decision support system can predict the future based on past observations. While not always entirely spot on, it can give a future estimation about aspects such as energy required, amounts of wastes produced etc. in the future. Therefore, processing plants can plan of time and adapt to specific needs. A human expert can provide this as well to some degree, but on a much smaller scale. Especially in time series forecasting, it is interesting to note that, even if a decision support model does not predict exact values, it is highly likely to predict trends of the value increasing or decreasing in certain ranges. In the next sections, we are going to describe the four machine learning models that were developed and which compose the Decision Support System of FENIX. Section 8.1 describes how we predict the quality of the extracted materials based on features such as temperature, extruder speed, etc. Section 8.2 describes the process of extracting heuristic rules based on existing results. Section 8.3 describes how FENIX provides time-series forecasting to predict the future of a variable based on past observations. Finally, Sect. 8.4 describes the process of classifying materials based on images.

scholarly journals Time series forecasting of agricultural product prices based on recurrent neural networks and its evaluation method

2020 ◽  
Vol 2 (8) ◽  
Author(s):  
Koichi Kurumatani
Keyword(s):  
Time Series ◽  
Probability Distribution ◽  
Time Series Forecasting ◽  
Agricultural Products ◽  
Superior Performance ◽  
Statistical Characteristics ◽  
Future Time ◽  
Forecasting Methods ◽  
The Future ◽  
Time Point

AbstractWe propose a time series forecasting method for the future prices of agricultural products and present the criteria by which forecasted future time series are evaluated in the context of statistical characteristics. Time series forecasting of agricultural products has the basic importance in maintaining the sustainability of agricultural production. The prices of agricultural products show seasonality in their time series, and conventional methods such as the auto-regressive integrated moving average (ARIMA or the Box Jenkins method) have tried to exploit this feature for forecasting. We expect that recurrent neural networks, representing the latest machine learning technology, can forecast future time series better than conventional methods. The measures used in evaluating the forecasted results are also of importance. In literature, the accuracy determined by the error rate at a specific time point in the future, is widely used for evaluation. We predict that, in addition to the error rate, the criterion for conservation of the statistical characteristics of the probability distribution function from the original past time series to the future time series in the forecasted future is also important. This is because some time series have a non-Gaussian probability distribution (such as the Lévy stable distribution) as a characteristic of the target system; for example, market prices on typical days change slightly, however on certain occasions, change dramatically. We implemented two methods for time series forecasting based on recurrent neutral network (RNN), one of which is called time-alignment of time point forecast (TATP), and another one is called direct future time series forecast (DFTS). They were evaluated using the two aforementioned criteria consisting of the accuracy and the conservation of the statistical characteristics of the probability distribution function. We found that after intensive training, TATP of LTSM shows superior performance in not only accuracy, but also the conservation compared to TATP of other RNNs. In DFTS, DFTS of LSTM cannot show the best performance in accuracy in RMS sense, but it shows superior performance in other criteria. The results suggest that the selection of forecasting methods depends on the evaluation criteria and that combinations of forecasting methods is useful based on the application. The advantage of our method is that the required length of time series for training is enough short, namely, we can forecast the whole cycle of future time series after training with even less than the half of the cycle, and it can be applied to the field where enough numbers of continuous data are not available.

scholarly journals Intelligent financial time series forecasting: A complex neuro-fuzzy approach with multi-swarm intelligence

2012 ◽  
Vol 22 (4) ◽  
pp. 787-800 ◽  
Author(s):  
Chunshien Li ◽  
Tai-Wei Chiang
Keyword(s):  
Time Series ◽  
Financial Time Series ◽  
Time Series Forecasting ◽  
Recursive Least Squares ◽  
Intelligent Computing ◽  
Nonlinear Functional ◽  
Financial Time ◽  
Computing Paradigm ◽  
Neuro Fuzzy ◽  
The Future

Financial investors often face an urgent need to predict the future. Accurate forecasting may allow investors to be aware of changes in financial markets in the future, so that they can reduce the risk of investment. In this paper, we present an intelligent computing paradigm, called the Complex Neuro-Fuzzy System (CNFS), applied to the problem of financial time series forecasting. The CNFS is an adaptive system, which is designed using Complex Fuzzy Sets (CFSs) whose membership functions are complex-valued and characterized within the unit disc of the complex plane. The application of CFSs to the CNFS can augment the adaptive capability of nonlinear functional mapping, which is valuable for nonlinear forecasting. Moreover, to optimize the CNFS for accurate forecasting, we devised a new hybrid learning method, called the HMSPSO-RLSE, which integrates in a hybrid way the so-called Hierarchical Multi-Swarm PSO (HMSPSO) and the wellknown Recursive Least Squares Estimator (RLSE). Three examples of financial time series are used to test the proposed approach, whose experimental results outperform those of other methods.

Deep Learning for text in limted data settings

2020 ◽  
Author(s):  
Pathikkumar Patel ◽  
Bhargav Lad ◽  
Jinan Fiaidhi
Keyword(s):  
Machine Learning ◽  
Time Series ◽  
Deep Learning ◽  
Sentiment Analysis ◽  
Transfer Learning ◽  
Text Classification ◽  
State Of The Art ◽  
Time Series Forecasting ◽  
Text Data ◽  
Performance Levels

During the last few years, RNN models have been extensively used and they have proven to be better for sequence and text data. RNNs have achieved state-of-the-art performance levels in several applications such as text classification, sequence to sequence modelling and time series forecasting. In this article we will review different Machine Learning and Deep Learning based approaches for text data and look at the results obtained from these methods. This work also explores the use of transfer learning in NLP and how it affects the performance of models on a specific application of sentiment analysis.

Sign in / Sign up

Export Citation Format

Share Document