Adjusting the type of neural networks to evaluate children's gait
Fragment książki (Rozdział w monografii)
MNiSW
20
Poziom I
| Status: | |
| Autorzy: | Skublewska-Paszkowska Maria, Powroźnik Paweł, Łukasik Edyta, Smołka Jakub, Miłosz Marek, Taczała Jolanta, Zdzienicka-Chyła Agnieszka, Kosiecz Anna |
| Dyscypliny: | |
| Aby zobaczyć szczegóły należy się zalogować. | |
| Wersja dokumentu: | Drukowana | Elektroniczna |
| Arkusze wydawnicze: | 0,8 |
| Język: | angielski |
| Strony: | 119 - 131 |
| Efekt badań statutowych | NIE |
| Materiał konferencyjny: | NIE |
| Publikacja OA: | TAK |
| Licencja: | |
| Sposób udostępnienia: | Witryna wydawcy |
| Wersja tekstu: | Ostateczna wersja opublikowana |
| Czas opublikowania: | W momencie opublikowania |
| Data opublikowania w OA: | 7 grudnia 2020 |
| Abstrakty: | angielski |
| Artificial neural networks are an excellent tool to classify patterns. Depending on the knowledge about dependencies between input and output data, different neural structures are used. The aim of the paper is to select the neural network structure, parameters and learning method to distinguish the gait of healthy children and those with disorders. The input vector to the neural network is: step length, step velocity, step duration, step width, step cadence, the maximal height of heel and toe during the step and average ankle angle. These items were generated from three-dimensional data recorded using a motion capture system (Vicon, Oxford Metrics Ltd., UK) and the biomechanical model Plug-in Gait. The children were walking along a straight 2.5 m path.Possessing the data describing the children's walk is a condition insufficient to be able to clearly assign a specific patient to groups of people characterised by a correct or incorrect walk. It is also desirable to separate individual groups. In other words,it should be shown that it is possible to extract a set of parameters from the data sets that will allow for an unambiguous classification of the children's gait. During the research, 1608 sets of samples suitable for classification and derived from both children with walking defects and healthy children were obtained. One way neural networks as well as the Support Vector Machine were tested as a classifier. The separability of the collected data using the tSNE method was also shown. |
