Delay Prediction

A new Markov-based method for real time prediction of network transmission time delays is introduced. The method considers a Multi-Layer Perceptron (MLP) neural model for the transmission network, where the number of neurons in the input layer is minimized so that the required calculations are reduced and the method can be implemented in the real-time. For this purpose, the Markov process order is estimated offline, using pr-recorded network time delay history. Unlike most of the previously existing methods, the proposed approach is both accurate and fast enough for a real time implementation. Using such a scheme for real-time estimation of the upcoming time delays, a variable state feedback gain control scheme is also proposed and applied to the predicted discretized model of the plant. The proposed approach is shown, through wellknown benchmark problems, to be both accurate and fast enough for a real time implementation.

This study aims to identify the main factors leading to charter flight departure delay through data mining. The data sample analysed consists of 5,484 flights operated by a European airline between 2014 and 2017. The tuned dataset of 33 features was used for modelling departure delay (e.g., if the flight delayed more than 15 minutes). The results proved the value of the proposed approach by an area under the receiver operating characteristic curve of 0.831 and supported knowledge extraction through the databased sensitivity analysis. The features related to previous flight delay information were considered as being the most influential toward current flight being delayed or not, which is consistent with the propagating effect of flight delays. However, it is not the reason for the previous delay nor the delay duration that accounted for the most relevance. Instead, a computed feature indicating if there were two or more registered reasons accounted for 33% of relevance. The contributions include also using a broader data mining approach supported by an extensive data understanding and preparation stage using both proprietary and open access data sources to build a comprehensive dataset.

Current train delay (TD) prediction systems do not take advantage of state-of-the-art tools and techniques for handling and extracting useful and actionable information from the large amount of endogenous (i.e., generated by the railway system itself) and exogenous (i.e., related to railway operation but generated by external phenomena) data available. Additionally, they are not designed in order to deal with the intrinsic time varying nature of the problem (e.g., regular changes in the nominal timetable, etc.). The purpose of this paper is to build a dynamic data-driven TD prediction system that exploits the most recent tools and techniques in the field of time varying big data analysis. In particular, we map the TD prediction problem into a time varying multivariate regression problem that allows exploiting both historical data about the train movements and exogenous data about the weather provided by the national weather services. The performance of these methods have been tuned through the state-of-the-art thresholdout technique, a very powerful procedure which relies on the differential privacy theory. Finally, the performance of two efficient implementations of shallow and deep extreme learning machines that fully exploit the recent in-memory large-scale data processing technologies have been compared with the current state-of-the-art TD prediction systems. Results on real-world data coming from the Italian railway network show that the proposal of this paper is able to remarkably improve the state-of-the-art systems.

State-of-the-art train delay prediction systems do not exploit historical train movements data collected by the railway information systems, but they rely on static rules built by expert of the railway infrastructure based on classical univariate statistic. The purpose of this paper is to build a data-driven train delay prediction system for large-scale railway networks which exploits the most recent Big Data technologies and learning algorithms. In particular, we propose a fast learning algorithm for predicting train delays based on the Extreme Learning Machine that fully exploits the recent in-memory large-scale data processing technologies. Our system is able to rapidly extract nontrivial information from the large amount of data available in order to make accurate predictions about different future states of the railway network. Results on real world data coming from the Italian railway network show that our proposal is able to improve the current state-of-the-art train delay prediction systems.

Technologies such as RFID and Bluetooth have received considerable attention for tracking indoor moving objects. In a time-critical indoor tracking scenario such as airport baggage handling, a bag has to move through a sequence of locations until it is loaded into the aircraft. Inefficiency or inaccuracy at any step can make the bag risky, i.e., the bag may be delayed at the airport or sent to a wrong airport. In this paper, we propose a novel probabilistic approach for predicting the risk of an indoor moving object in real-time. We propose a probabilistic flow graph (PFG) and an aggregated probabilistic flow graph (APFG) that capture the historical object transitions and the durations of the transitions. In the graphs, the probabilistic information is stored in a set of histograms. Then we use the flow graphs for obtaining a risk score of an online object and use it for predicting its riskiness. The paper reports a comprehensive experimental study with multiple synthetic data sets and a real baggage tracking data set. The experimental results show that the proposed method can identify the risky objects very accurately when they approach the bottleneck locations on their paths and can significantly reduce the operation cost.

This study aims to identify the main factors leading to charter flight departure delay through data mining. The data sample analysed consists of 5484 flights operated by a European airline between 2014 and 2017. The tuned dataset of 33 features was used for modelling departure delay (e.g., if the flight delayed > 15 min). The results proved the value of the proposed approach by an area under the receiver operating characteristic curve of 0.831 and supported knowledge extraction through the data-based sensitivity analysis. The features related to previous flight delay information were considered as being the most influential toward current flight being delayed or not, which is consistent with the propagating effect of flight delays. However, it is not the reason for the previous delay nor the delay duration that accounted for the most relevance. Instead, a computed feature indicating if there were two or more registered reasons accounted for 33% of relevance. The contributions include also using a broader data mining approach supported by an extensive data understanding and preparation stage using both proprietary and open access data sources to build a comprehensive dataset.