Statistics & Probability

Robotics is not possible without Mathematics and in that role of statistics and probability is not separable. We are living in the world where lots of data is moving around us. Today robots, self driving vehicle technologies, etc are equipped with various kind and number of sensors which provide large quantity of data. In order to infer/extract the required information from such data, statistics plays a vital role. The part of artificial intelligence like machine learning and deep learning are totally based on data. But in my experience without sufficient knowledge of statistics and probability, it is difficult to dive deep into these areas. One of the parts of my daily Job since few years is to validate the performance of automotive sensor(s). Hence use of statistics in an integral part of my Job. Plotting scatter plots, histograms, finding correlated data, evaluating probability of sensor performance in detecting objects, etc. all are part of statistics.

Similarly, field of robotics works on very basic principle of uncertainty. All the sensors provides measurement with some uncertainties, all the actuators works with uncertainties, no device is 100% accurate and hence robot perception, navigation and control, everything is dependent on uncertainties and hence on probability. Most of the available algorithms in the field somewhere uses probability to define the data and to work on that data.

In short, this branch of mathematics is indispensable in the field of Robotics and Automation. During my learning towards artificial intelligence and during my daily Job, I have gone through many concepts. Hence some of the good references I have shared below.


References

1. Introduction to Statistics by Sebastian Thrun – online free course on Udacity

  • Visualization Relationships in data
  • Probability
  • Estimation
  • Outlier and normal distribution
  • Binomial distribution
  • Central Limit Theorem
  • Inference
  • Regression and correlation