Working on a run-through of Codemy's pytorch5 and 6, essentially Machine
Learning on the Iris dataset. So far:
.png)
.png)
.png)
R.png)
.png)
.png)
.png)
Codemy is starting a Thursday Series, on using pygame. I'm here forit.
It's pygame as usual. One thing that is different, the update function has been
replaced by display.flip(), which is an optimized version that updates the entire
screen.
.png)
During World War 2, Ukraine was occupied by **Nazi Germany**. The German-occupied territory of Ukraine was divided between two new German administrative units: the District of Galicia of the Nazi General Government and the Reichskommissariat Ukraine¹.
While some Ukrainians chose to resist and fight against the German occupation forces by joining the Red Army or irregular partisan units conducting guerrilla warfare against the Germans, most Ukrainians, especially in western Ukraine, had little loyalty toward the Soviet Union. This was due to the Soviet Union's repressive occupation of eastern Ukraine in the interwar years and its role in overseeing a famine in the early 1930s called the Holodomor, which resulted in the deaths of millions of Ukrainians¹.
Some Ukrainians worked with or for the Nazis against the Allied forces. Ukrainian nationalists hoped that enthusiastic collaboration would enable them to re-establish an independent state. However, their collaboration with Nazi Germany involved a series of war crimes and crimes against humanity, including participation in the Holocaust in Ukraine and massacres of Poles in Volhynia and Eastern Galicia¹.
It is important to note that not all Ukrainians collaborated with Nazi Germany. Many Ukrainians fought on both sides during World War 2. The majority of ethnic Ukrainians, approximately 4.5 million, fought in the Red Army against the Germans². Others joined Communist partisans, including prominent commander Sydir Kovpak².
For more information on Ukrainian collaboration with Nazi Germany during World War 2, you can refer to this [Wikipedia article](^1^).
> "Ukrainian collaboration with Nazi Germany took place during the occupation of Poland and the Ukrainian SSR, USSR, by Nazi Germany during the Second World War." ¹
> "Ukrainians fought on both sides in the Second World War. By far the majority of ethnic Ukrainians, about 4.5 million, fought in the Red Army against the Germans." ²
> "In (Pres) Putin’s telling, the modern Ukrainian independence movement began not in 1917 but during World War II. Under the German occupation of Ukraine, between 1941 and 1944, some Ukrainian independence fighters aligned themselves with the Nazis, whom they viewed as saviors from Soviet oppression." ³
Source: Conversation with Bing, 9/29/2023
(1) Ukrainian collaboration with Nazi Germany - Wikipedia. https://en.wikipedia.org/wiki/Ukrainian_collaboration_with_Nazi_Germany.
(2) Second World War - Encyclopedia of Ukraine. https://www.encyclopediaofukraine.com/display.asp?linkpath=pages%5CS%5CE%5CSecondWorldWar.htm.
(3) Ukrainian collaboration with Nazi Germany - Wikipedia. https://en.wikipedia.org/wiki/Ukrainian_collaboration_with_Nazi_Germany.
(4) The 20th-Century History Behind Russia’s Invasion of Ukraine. https://www.smithsonianmag.com/history/the-20th-century-history-behind-russias-invasion-of-ukraine-180979672/.
(5) Explainer: What was the Nazi's First Ukrainian Division? | CTV News. https://www.ctvnews.ca/canada/how-was-veteran-yaroslav-hunka-s-military-unit-linked-to-the-nazis-1.6578579.
Working through pandas 26, 27, 28:
So what is actually going on here. As one looks at the original diabetes dataset
as a whole, it soons becomes obvious one doesn't really know how the y values
were arrived at. All we have to go on is our complex data. That is a machine learning situation.
For the fun of it, we will use 80% of the dataset to train the computer. We will then
ask it to provide y or target values for the 20% we have set aside. We will then evaluate
te computer's performance in this task.
We are asking the computer to build a linear regression model. That is the 'fit'. We
will also use linear regression to evaluate the predicted outcomes against the real
in our 20% of the data.
We can redo the exercise any number of times. We seem to be scoring with approximately
45 % accuracy every time...
This is the end to the pandas for machine learning series. We are going back to Machine
Learning, but this time with Pytorch.
* * *
.png)
.png)
.png)
.png)
.png)
Had a moment, this morning, reading how Chatgpt was being fine-tuned by
the work of humans; remote work that is. The latter is indeed a flourishing
market for many countries.
The Secret Ingredient of ChatGPT Is Human Advice - The New York Times (nytimes.com)
Managed to get the real obesity dataset in pandas. It's fun to look at the data, and
I don't anticipate any differences with the normalized data, but we'll just see...
.png)
.png)
Was reading, this morning, about neutrino research in Ontario. sounds clever.
Scientists are closing in on why the universe exists | CBC Radio
Ended up looking up spin1/2, which is a property of leptons. Finally found
a comprehensible account of it thanks to Chatgpt:
.PNG)
.PNG)
The last three tutorials in Codemy's Pandas for Machine Learning series involves
working with the diabetes dataset. Seems all very interesting at first glance, but then
not so fast. The data has been normalized!!
Model-based and sequential feature selection — scikit-learn 1.3.1 documentation
* * *
The **sklearn diabetes dataset** is a popular dataset used in machine learning. It consists of **442 samples** with **10 features** ¹². The dataset is often used to develop and test machine learning algorithms. Each sample has **10 different attributes**, such as age, sex, body mass index (BMI), average blood pressure, and six blood serum measurements ¹. The target variable is a quantitative measure of disease progression one year after baseline ¹.
.png)
.png)
.png)
.png)
.png)
.png)
.png)
.png)
.png)
.png)
.png)
.png)
.png)
.png)
.png)