“Ronald Fisher and Most Chance Estimation” is a presentation that I gave to the Bay Space Entrepreneurs in Statistics (BAES) on August 20, 2016 at Symation in Richmond, CA — simply north of Berkeley. BAES is a Meetup Group for folks within the utility of statistics to entrepreneurial ventures.
A video and slides from the presentation at the moment are on-line:
Ronald Fisher and Most Chance Estimation BAES Presentation
Most Chance Estimation (MLE) is likely one of the main foundational strategies of parameter estimation and statistical inference. It’s utilized in many fields from experimental particle physics, the place it’s commonly used to detect and measure the parameters of latest particles such because the Higgs Particle on the Massive Hadron Collider (LHC), to speech recognition the place it varieties the idea of the standard Hidden Markov Mannequin (HMM) primarily based speech recognition algorithms utilized by the Carnegie Mellon College (CMU) Sphinx open supply speech recognition engine and Nuance’s Dragon Naturally Talking.
Along with parameter estimation, MLE can be utilized for classification as is finished in speech recognition. Is that this utterance “I scream” or “ice cream” for instance? This discuss will focus on the origins of Most Chance Estimation within the pioneering work of Ronald Fisher, some historical past of the event and use of the tactic, and numerous sensible and theoretical issues that bedevil this fashionable, highly effective, however troublesome to make use of approach together with: vulnerability to outliers and the issue of robustness, sensible issues with becoming multidimensional fashions, correct normalization of the fashions, and restricted laptop energy.
The slides for this discuss can be discovered right here:
NOTE: For privateness causes, a nonetheless body is overlaid on the decrease left nook of the video. It’s possible you’ll discover you’ll be able to see by means of the shoulder of an viewers member; that is why.