Machine learning (ML) methods facilitate automated data mining. The authors compare the effectiveness of selected ML methods (RBF networks, Kohonen networks, and random forest) as modelling tools supporting the selection of materials in ecodesign. Applied in the design process, ML methods help benefit from the knowledge, experience and creativity of designers stored in historical data in databases. Implemented into a decision support system, the knowledge can be utilized – in the case under analysis – in the process of design of environmentally friendly products. The study was initiated with an analysis of input data for the selection of materials. The input data, specified in cooperation with designers, include both technological and environmental parameters which guarantee the desired compatibility of materials. Next, models were developed using selected ML methods. The models were assessed and implemented into an expert system. The authors show which models best fit their purpose and why. Models supporting the selection of materials, connections and disassembly methods help boost the recycling properties of designed products.

The paper examines the usage of Convolutional Bidirectional Recurrent Neural Network (CBRNN) for a problem of quality measurement in a music content. The key contribution in this approach, compared to the existing research, is that the examined model is evaluated in terms of detecting acoustic anomalies without the requirement to provide a reference (clean) signal. Since real music content may include some modes of instrumental sounds, speech and singing voice or different audio effects, it is more complex to analyze than clean speech or artificial signals, especially without a comparison to the known reference content. The presented results might be treated as a proof of concept, since some specific types of artefacts are covered in this paper (examples of quantization defect, missing sound, distortion of gain characteristics, extra noise sound). However, the described model can be easily expanded to detect other impairments or used as a pre-trained model for other transfer learning processes. To examine the model efficiency several experiments have been performed and reported in the paper. The raw audio samples were transformed into Mel-scaled spectrograms and transferred as input to the model, first independently, then along with additional features (Zero Crossing Rate, Spectral Contrast). According to the obtained results, there is a significant increase in overall accuracy (by 10.1%), if Spectral Contrast information is provided together with Mel-scaled spectrograms. The paper examines also the influence of recursive layers on effectiveness of the artefact classification task.