Your data sets are multi dimensional, unstructured or simply very large: extracting their substance is not a trivial task. I can help you decipher them and include the processing in your normal workflow or your production steps.
There are several ways to consider this problem:
- Modeling:
You have technical knowledge of your system and want to use it. I have experience with both complex thermodynamical modeling and simple production quality checking.- Example 1: you know how to measure the quality of your product based on several metrics, but it’s now done manually on excel by your technicians. I can help you automatize it and go industry 4.0!
- Example 2: based on your own work or literature, you have found a mathematical model to understand your samples. You now want to apply this model to the next million samples without going through the trouble of programing it.
- Prediction (machine learning):
You know what you want to get, but obtaining this information is too expensive. Depending on whether you need to understand causes and effects, I can help you get the right machine learning model.- Example 1: you have set up several sensors and metrics on your production line. You sample part of your production for an expensive quality control process. We can optimize this sampling.
- Example 2: you have categorized hundreds of images, but this process is time consuming. A predictive model can help you categorize the next thousand images automatically.
- Clustering:
You have large data sets and need to cluster in different categories your samples, without previous knowledge of these clusters. A typical application for this is statistical process control, where you want to detect defective samples and categorize these defects. Added to a control chart, you have a timely view of your processes.
