Applied Statistics - Week 7
Monday the 6th - Friday the 10th of January 2025
The following is a description of what we will go through during this
week of the course. The chapter references and computer exercises are
considered read, understood, and solved by the beginning of the
following class, where I'll shortly go through the exercise
solution.
General notes, links, and comments:
If last week got you interested in the intersection of statistics and machine learning, then perhaps take a look at some of the talks at the PHYSTAT - Statistics meets ML conference (which I unfortunately could not attend).
Monday:
I will go through the curriculum, and answer all the possible questions you may have,
perhaps sharpened by the problem set handing in.
In the exercises, we'll try a simple example of doing integration in many dimensions using simple simulation.
First, it is the estimate of pi, followed by the rational numbers in front of (hyper) volumes of balls
in many dimensions!
Reading:
You should at this point have read essentially the whole curriculum.
Lecture(s):
Stratification - beating the sqrt(N) law!
Afterwards, I'll essentially give very short versions of the last seven week's curriculum, with focus on what you want to hear about.
Computer Exercise(s):
Pi estimate with simulation: PiEstimate_original.ipynb
Tuesday:
The subject of the day will be advanced fitting, moving towards more complicated cases.
As stated, fitting is a bit of an artform, and there is little literature on the
subject - only (bitter?) experience! I've tried my best in the reading list below.
Reading:
Possibly Barlow page 184, section 10.2.2.
Possibly Cowan page 65.
Possibly Bevington chapters 6-8 (best of the three!).
Curve Fitting (Wikipedia)
Lecture(s):
Advanced Fitting
Computer Exercise(s):
Advanced fitting:
FittingTricks_1.XfarFromZero_original.ipynb
FittingTricks_2.CombinedParameters_original.ipynb
FittingTricks_3.HighCorrelations_original.ipynb
Friday:
The day will focus on time series, which is a separate subject, yet fairly straight forward,
once you get the hang of the idea. The associated exercise is inspired by biophysics.
Reading:
No reading - logic and reason suffices.
Lecture(s):
Time Series (Mathias)
Computer Exercise(s):
Time series: TimeSeries_original.ipynb
Last updated: 1st of January 2025.