(173k) Machine Learning Analysis of Multimodal Data from a Smartphone-Based Electrochemiluminescence Sensor.

Machine learning was applied for the data analysis of multimodal data from a smartphone based electrochemiluminescence (ECL) sensor. This portable sensor with low-cost alternative technology was used for the detection and quantification of phenolic compounds. The effective quenching of the ECL reaction by the phenolic compounds, vanillic acid and p-coumaric acid, was recorded by the signals of light intensity and electric current using apps of the sensor. Due to common problems present in sensor data such as non-linearity, multimodality, sensor-to-sensor variations, presence of anomalies, and ambiguity in key features, several machine learning strategies were explored. In contrast to the traditional calibration approach of extracting predetermined key features, the ML methods such as tri-layer neural net or boosted trees carried out effective regression tasks by learning higher patterns without pre-processing the key features. Combined multimodal features made 80% enhanced performance with multilayer neural net algorithms than the traditional approaches. The results demonstrated that the ML methods could provide a robust analysis framework for sensor data with noises and variability without preprocessing to extract features or examine ambiguous anomalies.