The graduate research project of a microfluidics particle-analyzing device based on impedance sensing
With the proposed device, we collected different signal data with unknown types of beads and cells; our task is to use machine learning algorithm to classify or make clusters of different beads based on their properties and signals.
- First approach: naive-bayes generative modeling with bivariate guassian models.
- Second approach: Bayesian unsupervised learning: Guassian Mixed models.
- Third approach: K-means algorithm
- Four approach: Deep neural network by pytorch