Tag Archives: learning

stanford’s ai class final feedback

Posted on by
3

I just completed the “final” for Stanford’s free online course – ai-class.  Ironic to post this the day after Stanford dropped their bid to come to New York.  It was taught by two Stanford professors – Peter Norvig and Sebastian Thrun.  I’ve sprinked my impressions througout the description.

How I found out

I found out from a student on the Stuyvesant high school robotics team.  There was also an article in the New York Times in August.

Why I took it

The topic sounded interesting.  I went to graduate school online and participate in the online forum coderanch.com.  I know that I enjoy learning while having the opportunity to discuss (or read discussions) from my classmates.  And I was very curious how the experiment in a massively distributed class would work out!  It went well.

How it worked

Area What I liked What could be better next time
Lectures 1-4 minute video segments.  Makes it easy to watch partial lectures at a time.  Great for busy people! I live in a good bandwidth area.  If I didn’t (or was traveling) a downloaded version or text dump would help.
Quizzes I loved the interactivity as you went.  When I was in college, I actively participated in lectures and it helped me learn.  I missed that in online classes and you completely nailed it!  I have no problem with asking things that haven’t been covered yet – it makes you think.  I also liked how later in the course, the professors started sharing if a quiz was meant to be easy or hard so you knew if you were overthinking. The only thing that bothered me on the quizzes was the tab order being broken, but this was fixed.
Homework Built well with what was learned in class.  It helped me realize what I didn’t understand as fully and also to understand better. Would have liked some more programming/application.
Midterm I liked that it could be printed because I was out all day Saturday and was able to do part of the exam on the go.  I liked that it was over a weekend since there was less than a week to work on it.  Effort != duration.  I don’t reliably have 3-4 hours during the week to do it. It was easier than I expected.
Final I liked that the questions were harder than the midterm in that they required more understanding. On the midterm, I found the PDF to be equivalent.  On the final, I felt like there were some tips/pointers in the online version that were not represented in the PDF.
Programming The two programming assignments was fun. I also did some programming to calculate values, check answers and just play with some of the concepts. It would have been nice to have more of these.  Or even link to the programming assignments for the “real” class.
Forums There were two forums: reddit and aiqus.  I started on reddit because I already used reddit for other things.  I then switeched to aiqus when links for each lecture/question started appearing on ai-class.com. Making aiqus the official forum earlier would have helped.  While it was well moderated, I feel like it was running into a limitation of the stack exchange (I presume) software.  In particular, during the exams, the large # of closed questions drowned out discussion.  I would have also liked to see some more discussions like we had in grad school rather than “how do I solve #1.”  One way to faciliate that is to have an official question of the week linked to from the course.  For example, “how could we apply X to Y” or “discuss stanley; the robotic car”.
Progress Bar The stie has a tab where you can see your progress.  You could see your score on the quiz/homework/exam along with what % of the week’s lecture you have completed. The biggest thing I would change is to provide a list of what you get wrong. (I was informed you can expand the score to see this. I guess minor usability thing since I missed it). I want to know which questions I got wrong way more than I want to know my score.  I had to rewatch each homework question to find the things I got wrong so I could learn why.
Announcements The home page had announcements like corrections and due dates.  Having due dates prominent helps.  And even though scores don’t matter, the course is heavily cumulative so you can’t really fall behind.  Plus discussions vary on the progress of the course. Some corrections were posted on the home page and some were not.  Others were quietly made on twitter/facebook/in the question itself.  Would be nice to have this centralized.
Office hours I didn’t participate. I can’t possibly have feedback on something I didn’t look at!

My measure of success

I know a lot of things I didn’t before the class started.  That is *much* more important than the score.  While I did buy a used copy of the 2nd edition of the textbook, the class itself was free (and buying the book was optional.)  When I started, I figured I would try it and see if I wanted to stay with it.  Every week I learned a lot and it was enjoyable so I stuck with it. I enjoyed learning the theory.  In particular liked how it applied to robotics and language processing.

How I did

I’d like to repeat the part about learning being the important part here.  When one says that someone often chimes in “oh, that’s just because you didn’t do well.”  So I’ll share that my average is in the mid 90s.  (Not including the final which isn’t due yet.)

My comments to the upset people

  1. If you got something wrong, don’t make excuses. Getting it wrong is about learning. Even if it about communication.  Even if the “communication error” was at the Stanford end.  In the real world you communicate with people or unclear and ambiguous.  It’s a valuable life skill.
  2. Making mistakes it good.  Especially in a situation where it doesn’t matter.  I like to say that I’d rather make mistakes on CodeRanch than it work because the effects are smaller. Same thing here.  This doesn’t count towards your job or degree/GPA.  The only thing that matters is what you learn.  Getting something wrong makes it memorable.

What’s next?

They ended the class with a little humor such as a quiz question “did you understand?” (yes is the correct answer) and a homework question asking how Stanford did in the  2005 DARPAUrban Challenge (they won – an accomplishment the instructions should rightfully be proud of.)

A lot more courses are being offered next semester.  See the full list at class central.  I signed up for Human Computer Interaction and Software as  Service.  Two might be too much but I’m not clear on how much they overlap since the # weeks isn’t specified.

Other opinions

A lot of people wrote comments on the web.  My favorites:

  1. a perspective from Association for Learning Technology
  2. Seb Schmoller’s weekly comments

is two people pairing better in a training class?

Posted on by
Reply

This girl and her coworker are pairing together in this class. She said to the instructor pairing is pointless. Her partner in crime said its distracting and silly.

Someone e-mailed me this story suggesting it would make a good blog.  I feel for both sides of the story here so I’m going to present an argument for both points of view.  I bet the e-mailer didn’t see that coming!

Pair programming on a project

On an actual task, I think pair programming is great.  It is faster, higher quality, both people know the code and provides a great way to learn from each other.  In a $5,000 training class, I have mixed feelings.

Pros – why two students per computer in class rocks

I’ve never actually done this so it is more hypothetical.

  1. You get stuck less because there are two minds.
  2. You go down less dead ends completing the labs faster.
  3. Your partner can point out when you aren’t getting something or doing it the longer way.
  4. One person can explain to the other when it takes longer for one person to grasp a concept.
  5. You can learn tricks from the other person like how to use the IDE better.
  6. You don’t have to ask the instructor for help as often – which is a big benefit if a lot of people in the class have questions.

Cons – why two students per computer in class is a waste of time

  1. I want to make as many mistakes as humanly possible while in class so I don’t on the job.
  2. I want to get hands on experience *doing* each step rather than just watching.  I’m a tactile learner and I’m going to retain more by doing.
  3. I want to gain experience with debugging/troubleshooting in a controlled setting where the instructor can bail me out as needed.
  4. I think faster than the average person and always finish labs very early.  I want to use that time to explore the subject matter rather than have the whole lab take longer.
  5. Sometimes asking the instructor a question sparks a good discussion.
  6. Sometimes I want to try something that isn’t strictly covered in class so I can ask the instructor/expert when I get stuck.  Can’t do this with a partner.

The ideal in my mind

Having one computer for each student and asking the person next to you when you have a question.  Benefits:

  1. Frees up the instructor for the hard questions
  2. Lets you see the mistakes of two people rather than just your own
  3. Gain practice trying to explain things

What do you think?  Do you prefer pairing with someone else in class?   Ultimately, I think it depends on the learning style of both people.  For my style, having my own computer is more conducive to learning a technology.

Which Database to Start With?

Posted on by
14

When people ask me how to learn to use a database or how to write SQL queries, I tell them to pick a database system and immerse themselves in it. In fact that advice goes for a lot of software technologies: just immerse yourself in a language, as programming tutorials are easy to come by these days. On the other hand, when people ask me which database software to use, I tend to give pause. Most of the time, I recommend MySQL for beginners since it tends to be the most light-weight system to install and use, but I know it’s not often the easiest to understand. With the advent of new light-weight database editions of often heavier products, perhaps it’s time I reconsider the issue.

1. MySQL: Free, lightweight, and readily available

MySQL stands out as the easiest for users to start with, in part because most people can get access to a MySQL database without having to setup anything. Most, if not all, hosting companies that offer database support do so in the form of a MySQL database. The only disadvantage with hosting solutions is that users lose the ability to run local applications on the database, often relying on phpMyAdmin for all database changes. I recommend anyone serious about learning MySQL download and install it themselves, as there are plenty of installation platforms supported.

The good: Free. Easy to download and/or find an existing database to work with. Somewhat easy to install. Lots of free tools available. Good documentation.
The bad: If the installation or auto-configuration breaks, user is left spending hours diagnosing the problems. The MySQL GUI tools, while nice, have to be downloaded separately from the server. Limited support. Clustering and support of large transaction systems is not uncommon. Also, it can be buggy and unpredictable at times, as I’ve seen in practice.

2. Oracle: Heavy and Powerful

Oracle is one of the oldest database systems and stands out as a powerhouse among databases given its vast support for advanced clustering, memory management, and query optimization. If you need something robust, powerful, and able to support millions or billions of transactions a day, it’s the best there is. Oracle needs to be licensed for a production environment, although developers can download a free limited-use version which is good for building an application.

The good: Powerful. Can do some really cool things for those that appreciate it. Extremely scalable.
The bad: Often large and time-consuming installation. Least user friendly of all the database systems, although it’s gotten better over the last few years. Not free. Not a wide variety of tools, free or otherwise, to manipulate the database.

3. Microsoft SQL Server: Easy to use administration interface, often powerful

Microsoft SQL Server has matured greatly over the last 10 years into a decent rival of Oracle. I like MS SQL Server in that it hides a lot of the underlying configuration information from the user. On the other hand, I dislike MS SQL server in that it hides a lot of the underlying configuration information from the user. Double-edged sword, I know. Like Oracle, you need a license if you want to use it in a production environment.

The good: Easy to set up new databases and administer them. Best for those who have no idea how to administer a database. New express editions can be used for free.
The bad: Over-simplifies a lot for advanced users, making it harder to optimize. Not free. Developer edition has nominal cost, although it probably should be free.

Other Databases

This article is not meant to be the end-all for database software discussion, but a beginning guide of the big three database systems for those who are not well-versed in the area. To cover every possible database software, such as PostgreSQL or DB2, as well as countless others, would take a book or two. Most students starting out just need to find a single database and start ‘playing’ with it until they get the hang of it, rather than an exhaustive discussion of which database is best.

Non-standard Databases

Some of you may be more familiar with embedded databases such HSQLDB, SQLite, or Derby than the ones I have mentioned. Rarely do I see beginners using embedded databases, so perhaps I’ll write an article about such systems down the road. Also, I have not purposely not mentioned Microsoft Access as a learning database, simply because I don’t consider it standard database software, but rather a glorified Excel spreadsheet. Most of teaching someone how to use a regular database after using Access, is convincing them all databases are not like Access.

My favorite database? If I’m teaching or writing a relatively simple web-application, MySQL. If someone else is paying for the license and the application is large enough, Oracle.