The gradient decent approach is used in many algorithms to minimize loss functions. In this introduction we will see how exactly a gradient descent works. In addition, some special features will be pointed out. We will be guided by a practical example.…
Added by Frank Raulf on June 24, 2020 at 7:00am — No Comments
There are many ways to deal with time-data. Sometimes one can use it as time-series to take possible trends into account. Sometimes this is not possible because time can not be arranged in a sequence. For example, if there are just weekdays (1 to 7) in a dataset over several month. In this case one could use one-hot-encoding. However, considering minutes or seconds of a day one-hot-encoding might lead to high complexity. Another approach is to make time cyclical. This approach leads to a…Continue
Added by Frank Raulf on January 26, 2020 at 4:00am — No Comments
For decision making, human perception tends to arrange probabilities into above 50% and below - which is plausible. For most probabilistic models in contrast, this is not the case at all. Frequently, resulting probabilities are neither normal distributed between zero and one with a mean of 0.5 nor correct in terms of absolute values. This is not seldom an issue accompanied with the existence of a minority class - in the underlying dataset.
For example, if the result of a…Continue
Added by Frank Raulf on January 4, 2020 at 3:00am — No Comments
Bayesian inference is the re-allocation of credibilities over possibilities [Krutschke 2015]. This means that a bayesian statistician has an “a priori” opinion regarding the probabilities of an event:
By observing new data x, the statistician will adjust his opinions to get the "a posteriori" probabilities.
The conditional probability of an event d given x is the share of the joint…Continue
Added by Frank Raulf on January 3, 2020 at 4:30am — No Comments
This post is the third one of a series regarding loops in R an Python.
The first one was Different kinds of loops in R. The recommendation…Continue
The importance of completeness of linear regressions is an often-discussed issue. By leaving out relevant variables the coefficients might be inconsistent.
But why on earth?!
Assuming a linear complete model of the form:
z = a + bx + cy + ε.
Where z is supposed to be dependent, x and y are independent and ε is the error term.
Now we drop y to check…Continue
Added by Frank Raulf on November 13, 2019 at 2:00am — No Comments
The positive reactions on my last post: “Different kinds of loops in R” lead me to compare some different versions of loops in R, RCPP (C++ integration of R). To see a bigger picture, I apply the Python for-loop additionally. The comparison focuses on the runtime for non-costly tasks with different numbers of iterations. For comparison purpose I create vectors in the form of (R syntax):
Vector <- 1:k
k = (1.000, 100.000, 1.000.000)
The task is to…Continue
Normally, it is better to avoid loops in R. But for highly individual tasks a vectorization is not always possible. Hence, a loop is needed – if the problem is decomposable.
Which different kinds of loops exist in R and which one to use in which situation?
In each programming language, for- and while-loops (sometimes until-loops) exist. These loops are sequential and not that fast – in R.
Added by Frank Raulf on August 12, 2019 at 12:30am — No Comments