Montag, 28. Januar 2013

ENNDT / sPIqr

I took part in the beta testing session of Marco Ripa's new dynamically generated IQ test. From 29 participants, only 4 managed to solve at least one of the assignments. I solved 10 and apparently this was the highest number of items anyone solved in this session. In the night there was another session, in which I did not participate, and in that session one participant solved 13, including some of the hardest items. (There were 37 items in total. The time limit was 160 minutes.)

As a consequence Marco invited me to join the sPIqr society which he had founded. I gladly accepted. So now I am a member of a high IQ society which is 100 times more selective than Mensa, i.e. only 1 out of 100 (potential) Mensa members has an IQ high enough to join it. This makes it the most selective high IQ society I'm currently a member of. Some days ago I learned that every year about 90,000 babies are born in Austria, including 9,000 in Vienna. If the hypothesis of the Gaussian distribution of IQ scores is right, then I'm apparently among the top three of intelligent persons born in Vienna in 1983.

Montag, 21. Januar 2013

Nydegger Intelligence Test

So my current scores on this test are the following:

Abstract reasoning subtest standard score: 148 (Preliminary norm)
Spatial subtest standard score: 140 (Preliminary norm)
Numerical subtest standard score: 152 (Preliminary norm)
Verbal subtest standard score: 153 (Estimated norm)
Logical subtest standard score: 162 (Estimated norm)
____________________________________________________
NIT-Form I I.Q. score: 153 (Estimated)

I wrote "current scores" since there are only preliminary norms for three subtests yet, and the other scores are estimates. Compared to the point of time when there was just an estimated norm of the numerical subtest I gained 6 IQ points.

This total score, 153, and the numerical subtest score, 152, are very close to my score on Ivan Ivec's Numerus Basic test, on which I got 152.

Sonntag, 20. Januar 2013

Lerntechniken

Auf e-fellows.net wurde von einer Studie berichtet, wonach die meisten Lerntechniken, wie etwa Unterstreichen oder Zusammenfassen, keinen Nutzen haben. Das entspricht an sich meiner Erfahrung. Ich bin, nach 10 Jahren Studium, zum Schluss gekommen, dass es reicht, den Stoff einmal durchzulesen, und wenn man beim ersten Durchlesen schlampig war, dann halt die Passagen, die man noch nicht ordentlich gelernt hat, noch einmal. Natürlich kann es sein, dass man dann bei der Prüfung durchfällt, aber man glaube mir: die anderen Lerntechniken hätten zu keinem besseren Ergebnis geführt. Sollte man bei den Prüfungen öfters durchfallen, heißt das eben, dass man für das Studium nicht geeignet ist.

Donnerstag, 17. Januar 2013

Medizinstudium - kompakt

In den letzten zwei Tagen habe ich mit Interesse einige alte (vor zehn Jahren aktuelle) Lehrbücher aus meinem Medizinstudium gelesen und teilweise auch Wissenslücken aufgefrischt. Dabei ist mir eingefallen, dass erschreckenderweise alles, was ich am Medizinstudium interessant fand, in zwei Jahren abgehandelt werden könnte:

1. Semester
- Chemie
- Physik

2. Semester
- Biochemie
- Physiologie

3. Semester
- Funktionelle Pathologie
- Mikrobiologie

4. Semester
- Pharmakologie

Das könnte man auch zur Grundlage eines Studienplans machen. Die Zellbiologie und Genetik, die im alten Studienplan Medizin Teil des (von mir als obsolet erachteten) Prüfungsfachs Biologie waren, könnte man im Fach Biochemie zusätzlich zum bisherigen Stoff der Biochemie abhandeln. Anatomie und Histologie fand ich, bis auf die Grundlagen, unnötig, und es wäre kein Problem, diese Grundlagen im Rahmen einer Physiologievorlesung abzuhandeln. Die Pathologische Anatomie ist ein unglaublich umfang- und detailreiches Fach und das Wissen wird im Detail höchstens von angehenden Fachärzten für Pathologie gebraucht. Deshalb würde ich sie ersatzlos streichen, eine Vorlesung über die Funktionelle Pathologie reicht.

Im klinischen Abschnitt könnte man die Prüfungsordnung weitgehend beibehalten, aber Sozialmedizin, Gerichtsmedizin und Zahnmedizin könnte man abschaffen (auch wenn es sehr leichte Prüfungen waren) und die Neurologie und die Psychiatrie zu einem Fach vereinigen.

Samstag, 5. Januar 2013

My Master's thesis

My Master's thesis is soon going to be finished; therefore, I will write a couple of words about its making.

I started in July 2012. I had found the topic at the website of the Algorithms and Data Structures Group of our university and contacted the university assistant who had proposed it. We agreed that I would start by implementing a general framework, and then we would have a meeting to discuss the details.

In effect I implemented a working algorithm within two weeks. But this algorithm had to be enhanced by various variants, both exact and heuristic ones. I had to learn how to handle ILOG CPLEX, and the ultimate challenge was to implement a variant of the algorithm that makes use of the technique of column generation to keep the number of constraints to solve as low as possible.

In the end I succeeded at all tasks, and spent a lot of time waiting for my program to generate results for tons of test instances.

Writing the actual thesis was no problem for me and went rather fast. I also had to deliver a presentation about my thesis, and this presentation was well received by the audience (almost the entire ADS Group was listening to it), I got grade "excellent" on it.

Now there is vacation, and I hope that afterwards, the assistant will agree that the thesis is ready to be submitted. In that case I would be able to submit it by the end of January 2013, and in the beginning of March 2013 I would have my final exam and graduate from university.

So all in all I spent seven months since the start of the work on this thesis, which is just one month more than the time you are expected to finish your thesis in. Most students need longer than that. I had a lot of fun working on the thesis and I would certainly enjoy doing more of such "research" work.