In this paper, the problem of the identification of undesirable events is discussed. Such events can be poorly represented in the historical data, and it is predominantly impossible to learn from past examples. The discussed issue is considered in the work in the context of two use cases in which vibration and temperature measurements collected by wireless sensors are analysed. These use cases include crushers at a coal-fired power plant and gantries in a steelworks converter. The awareness, resulting from the cooperation with industry, of the need for a system that works in cold start conditions and does not flood the machine operator with alarms was the motivation for proposing a new predictive maintenance method. The proposed solution is based on the methods of outlier identification. These methods are applied to the collected data that was transformed into a multidimensional feature vector. The novelty of the proposed solution stems from the creation of a methodology for the reduction of false positive alarms, which was applied to a system identifying undesirable events. This methodology is based on the adaptation of the system to the analysed data, the interaction with the dispatcher, and the use of the XAI (eXplainable Artificial Intelligence) method. The experiments performed on several data sets showed that the proposed method reduced false alarms by 90.25% on average in relation to the performance of the stand-alone outlier detection method. The obtained results allowed for the implementation of the developed method to a system operating in a real industrial facility. The conducted research may be valuable for systems with a cold start problem where frequent alarms can lead to discouragement and disregard for the system by the user.

Coal mining requires working in hazardous conditions. Miners in an underground coal mine can face several threats, such as, e.g. methane explosions. To provide protection for people working underground, systems for active monitoring of production processes are typically used. One of their fundamental applications is screening dangerous gas concentrations (methane in particular) to prevent spontaneous explosions. Such a system is the source of the data set containing raw data collected at an underground coal mine. The data is collected from 28 different sensors placed at various locations around the coal mine. All the attributes except one are numeric, and the examples collected form a time series. This data set can be used in a variety of analytical tasks, including classification, regression, time series and stream data analysis.

In the DECODE project, data were collected from 3,114 surveys filled by symptomatic patients RT-qPCR tested for SARS-CoV-2 in a single university centre in March-September 2020. The population demonstrated balanced sex and age with 759 SARS-CoV-2( +) patients. The most discriminative symptoms in SARS-CoV-2( +) patients at early infection stage were loss of taste/smell (OR = 3.33,p < 0.0001), body temperature above 38℃ (OR = 1.67,p < 0.0001), muscle aches (OR = 1.30,p = 0.0242), headache (OR = 1.27,p = 0.0405), cough (OR = 1.26,p = 0.0477). Dyspnea was more often reported among SARS-CoV-2(-) (OR = 0.55,p < 0.0001). Cough and dyspnea were 3.5 times more frequent among SARS-CoV-2(-) (OR = 0.28,p < 0.0001). Co-occurrence of cough, muscle aches, headache, loss of taste/smell (OR = 4.72,p = 0.0015) appeared significant, although co-occurrence of two symptoms only, cough and loss of smell or taste, means OR = 2.49 (p < 0.0001). Temperature > 38℃ with cough was most frequent in men (20%), while loss of taste/smell with cough in women (17%). For younger people, taste/smell impairment is sufficient to characterise infection, whereas in older patients co-occurrence of fever and cough is necessary. The presented study objectifies the single symptoms and interactions significance in COVID-19 diagnoses and demonstrates diverse symptomatology in patient groups.