Optimal Adaptive Prediction Intervals for Electricity Load Forecasting in Distribution Systems via Reinforcement Learning

Prediction intervals (PIs) offer an effective tool for quantifying uncertainty of loads in distribution systems. The traditional central PIs cannot adapt well to skewed distributions, and their offline training fashion is vulnerable to the unforeseen change in future load patterns. Therefore, we propose an optimal PI estimation approach, which is online and adaptive to different data distributions by adaptively determining symmetric or asymmetric probability proportion pairs for quantiles of PIs’ bounds. It relies on the online learning ability of reinforcement learning (RL) to integrate the two online tasks, i.e., the adaptive selection of probability proportion pairs and quantile predictions, both of which are modeled by neural networks. As such, the quality of quantiles-formed PI can guide the selection process of optimal probability proportion pairs, which forms a closed loop to improve PIs’ quality. Furthermore, to improve the learning efficiency of quantile forecasts, a prioritized experience replay (PER) strategy is proposed for online quantile regression processes. Case studies on both load and net load demonstrate that the proposed method can better adapt to data distribution compared with online central PIs method. Compared with offline-trained methods, it obtains PIs with better quality and is more robust against concept drift.

Optimal Adaptive Prediction Intervals for Electricity Load Forecasting in Distribution Systems via Reinforcement Learning

Prediction intervals (PIs) offer an effective tool for quantifying uncertainty of loads in distribution systems. The traditional central PIs cannot adapt well to skewed distributions, and their offline training fashion is vulnerable to the unforeseen change in future load patterns. Therefore, we propose an optimal PI estimation approach, which is online and adaptive to different data distributions by adaptively determining symmetric or asymmetric probability proportion pairs for quantiles of PIs’ bounds. It relies on the online learning ability of reinforcement learning (RL) to integrate the two online tasks, i.e., the adaptive selection of probability proportion pairs and quantile predictions, both of which are modeled by neural networks. As such, the quality of quantiles-formed PI can guide the selection process of optimal probability proportion pairs, which forms a closed loop to improve PIs’ quality. Furthermore, to improve the learning efficiency of quantile forecasts, a prioritized experience replay (PER) strategy is proposed for online quantile regression processes. Case studies on both load and net load demonstrate that the proposed method can better adapt to data distribution compared with online central PIs method. Compared with offline-trained methods, it obtains PIs with better quality and is more robust against concept drift.

Strong Algorithms for the Ordinal Matroid Secretary Problem

In the ordinal matroid secretary problem (MSP), candidates do not reveal numerical weights, but the decision maker can still discern if a candidate is better than another. An algorithm [Formula: see text] is probability-competitive if every element from the optimum appears with probability [Formula: see text] in the output. This measure is stronger than the standard utility competitiveness. Our main result is the introduction of a technique based on forbidden sets to design algorithms with strong probability-competitive ratios on many matroid classes. We improve upon the guarantees for almost every matroid class considered in the MSP literature. In particular, we achieve probability-competitive ratios of 4 for graphic matroids and of [Formula: see text] for laminar matroids. Additionally, we modify Kleinberg’s utility-competitive algorithm for uniform matroids in order to obtain an asymptotically optimal probability-competitive algorithm. We also contribute algorithms for the ordinal MSP on arbitrary matroids.

System Maintenance Using Several Imperfect Repairs Before a Perfect Repair

Allowing several imperfect repairs before a perfect repair can lead to a highly reliable and efficient system by reducing repair time and repair cost. Assuming exponential lifetime and exponential repair time, we determine the optimal probability $p$ of choosing a perfect repair over an imperfect repair after each failure. Based on either the limiting availability or the limiting average repair cost per unit time, we determine the optimal number of imperfect repairs before conducting a perfect repair.

The poor prognosis and influencing factors of high D-dimer levels for COVID-19 patients

To explore the value, and influencing factors, of D-dimer on the prognosis of patients with COVID-19. A total of 1,114 patients with confirmed COVID-19 who were admitted to three designated COVID-19 hospitals in Wuhan, China from January 18, 2020, to March 24, 2020, were included in this study. We examined the relationship between peripheral blood levels of D-dimer, and clinical classification and prognosis, as well as its related influencing factors. D-dimer levels were found to be related to the clinical classification and the prognosis of clinical outcome. D-dimer levels were more likely to be abnormal in severely and critically ill patients compared with mild and ordinary cases, while D-dimer levels of patients who had died were significantly higher than those of surviving patients according to the results of the first and last lab tests. The results from ROC analyses for mortality risk showed that the AUCs of D-dimer were 0.909, YI was 0.765 at the last lab test, and a D-dimer value of 2.025 mg/L was regarded to be the optimal probability cutoff for a prognosis of death. In addition, we found that patients with advanced age, male gender, dyspnea symptoms, and some underlying diseases have a higher D-dimer value (p < 0.05). In short, D-dimer is related to the clinical classification and can be used to evaluate the prognosis of COVID-19 patients. The D-dimer value of 2.025 mg/L was the optimal probability cutoff for judging an outcome of death. Advanced age, male gender, dyspnea symptoms, and some underlying diseases are influencing factors for D-dimer levels, which impacts the prognosis of patients.

Predicting Care Homes with Limited Residential Information

IntroductionResearch involving care homes is often difficult due to a lack of data and ethical issues. Wales (United Kingdom) contains approximately 1.3million residences, of these 717 are officially recorded as care homes for older people. Objectives and ApproachOur objective was to develop a predictive methodology for identifying care homes in administrative data. We used two data sources within the Secure Anonymised Information Linkage Databank to conduct our study. The Welsh Demographic Service Dataset (WDSD) contains all residences in Wales and demographic details of their occupants. An anonymised dataset of deterministically matched care home addresses was used to determine which of the residences in the WDSD were care homes. We used details in the WDSD to determine the average age of the occupants, the number of people who moved into the residence in a year, and the number of people who died in a year. We were interested in care homes for older people and restricted all the residences in the WDSD to only those with an average age of 50+ years. We applied logistic regression to determine a probabilistic match for care homes based on the above characteristics. We determined an optimal cut-point for the probability of a residence being a care home based on the sensitivity and specificity. ResultsRestricting the WDSD to have an average age of occupants of 50+ created a dataset of 3,939 residences, containing 562 care homes. After applying the logistic model to predict the care homes, we found an optimal probability cut-point which resulted in 548 true positives, 105 false positives, 14 false negatives, and 3,272 true negatives. ImplicationsIdentification of care homes in an anonymised databank using only demographic data allows research into healthcare pathways for this hard to reach and under-researched population.

Globe-hopping

We consider versions of the grasshopper problem (Goulko & Kent 2017 Proc. R. Soc. A 473 , 20170494) on the circle and the sphere, which are relevant to Bell inequalities. For a circle of circumference 2 π , we show that for unconstrained lawns of any length and arbitrary jump lengths, the supremum of the probability for the grasshopper’s jump to stay on the lawn is one. For antipodal lawns, which by definition contain precisely one of each pair of opposite points and have length π , we show this is true except when the jump length ϕ is of the form π ( p / q ) with p , q coprime and p odd. For these jump lengths, we show the optimal probability is 1 − 1/ q and construct optimal lawns. For a pair of antipodal lawns, we show that the optimal probability of jumping from one onto the other is 1 − 1/ q for p , q coprime, p odd and q even, and one in all other cases. For an antipodal lawn on the sphere, it is known (Kent & Pitalúa-García 2014 Phys. Rev. A 90 , 062124) that if ϕ = π / q , where q ∈ N , then the optimal retention probability of 1 − 1/ q for the grasshopper’s jump is provided by a hemispherical lawn. We show that in all other cases where 0 < ϕ < π /2, hemispherical lawns are not optimal, disproving the hemispherical colouring maximality hypotheses (Kent & Pitalúa-García 2014 Phys. Rev. A 90 , 062124). We discuss the implications for Bell experiments and related cryptographic tests.

Application and Verification of the ECMWF Precipitation Type Forecast Product (PTYPE) in China

&lt;p&gt;The ECMWF (European Centre for Medium-Range Weather Forecasts) precipitation type forecast products&amp;#8212;PTYPE are verified using the weather observations of more than 2000 stations in China of the past three winter half years (October to next March). The products include the deterministic forecast from High-resolution model (HRE) and the probability forecast from ensemble prediction system (EPS). Based on the verification results, optimal probability thresholds approaches under criteria of TS maximization (TSmax), frequency match (Bias1) and HSS maximization (HSSmax) are used to improve the deterministic precipitation type forecast skill. The researched precipitation types include rain, sleet, snow and freezing rain.&lt;/p&gt;&lt;p&gt;The verification results show that the proportion correct of deterministic forecast of ECMWF high-resolution model is mostly larger than 90% and the TSs of rain and snow are high, next is freezing rain, and the TS of sleet is small indicating that the forecast skill of sleet is limited. The rain and snow separating line of deterministic forecasts show errors of a little south in short-range and more and more significant north following elongating lead times in medium-range. The area of sleet forecasts is smaller than observations and the freezing rain is bigger for the high-resolution deterministic forecast. The ensemble prediction system offsets these errors partly by probability forecast. The probability forecast of rain from the ensemble prediction system is smaller than the observation frequency and the probability forecast of snow is larger in short-range and smaller in medium-range than the observation frequency. However, there are some forecast skills for all of these probability forecasts. There are advantages of ensemble prediction system compared to the high-resolution deterministic model. For rain and snow, for some special cost/loss ratio events the EPS is better than the HRD. For sleet and freezing rain, the EPS is better than the HRD significantly, especially for the freezing rain.&lt;/p&gt;&lt;p&gt;The optimal thresholds of snow and freezing rain are largest which are about 50%~90%, decreasing with elongating lead times. The thresholds of rain are small which are about 10%~20%, increasing with elongating lead times. The thresholds of sleet are the smallest which are under 10%. The verifications show that the approach of optimal probability threshold based on EPS can improve the forecast skill of precipitation type. The proportion correct of HRD is about 92%. Bias1 and TSmax improve it and the improvement of HSSmax is the most significant which is about 94%. The HSS of HRD is about 0.77~0.65. Bias1 increases 0.02 and TSmax increases more. The improvement of HSSmax is the biggest which is about 0.81~0.68 and the increasing rate is around 4%. From the verifications of every kinds of precipitation types, it is demonstrated that the approach of optimal probability threshold improves the performance of rain and snow forecasts significantly compared to the HRD and decreases the forecast area and missing of freezing rain and sleet which are forecasted more areas and false alarms by the HRD.&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Key words: &lt;/strong&gt;ECMWF; ensemble prediction system&amp;#65307;precipitation type forecast; approach of optimal probability threshold; verification&lt;/p&gt;

Cryptanalysis of the Lightweight Block Cipher BORON

This paper provides security evaluations of a lightweight block cipher called BORON proposed by Bansod et al. There is no third-party cryptanalysis towards BORON. Designers only provided coarse and simple security analysis. To fill this gap, security bounds of BORON against differential and linear cryptanalysis are presented in this paper. By automatic models based on the SMT solver STP, we search for differential and linear trails with the minimal number of active S-boxes and trails with optimal probability and bias. Then, we present key-recovery attacks towards round-reduced BORON. This paper is the first third-party cryptanalysis towards BORON.