ID3 Decision trees ūüĆ≥

As part of the Machine Learning class in (CIS 678) at GVSU, Camila and I¬†created a decision tree tool that¬†analyzes and processes a data-set, where each entry has a different combination of attributes and an expected¬†outcome.¬†The system¬†then finds the ¬ębest¬Ľ attributes, that allow deciding which¬†outcome will occur with the least amount of¬†questions,¬†for previously seen or ¬†new data.

The website & tool takes¬†data-sets in the form of a simple¬†text-file, then parses & processes it and finally creates a decision tree ūüĆ≥ based on that data. Please check it out and let us¬†know what you think and how you like it.

The tool can be accessed on this sub-site.


Tree example

DailyUI #002 – Credit Card Checkout

Today’s assignment was to create a credit card checkout.

To give myself a better context, I decided to checkout a monitor and took an existing online-shop to help guide¬†some of my¬†decisions (you can guess which one it is ūüėČ ).¬†Please let me know what you like and what you think needs improvement.

Day 002 - Credit Card Checkout

DailyUI #001 – Sign Up Screen

Hello everyone,
Today I completed my first DailyUI-Challenge #001. The task was to create a Sign Up screen.

My idea was to make it as simple as possible. Consequently I combined Sign Up and Log In. First it would start by letting the user add their email address. Then, if they already have an account it would then prompt them to enter their password. Since the user does not have an account yet, a confirmation-mail will be sent to him, to verify that the provided address actually is controlled by him.

Being able to show, whether a user needs to Sign Up or Log In, implies that the website discloses, if an email account already exists. While this brings certain drawbacks it might be okay, depending on your website.
Please let me know how you like it!


How about fixing Blackboard?

One of the more popular online learning (and training) management systems for high schools¬†and universities in the United States is Blackboard. However,¬†despite the fact that¬†it has a lot to offer, many of it’s features are hidden deep in the interface.

We all came across¬†it [ or at least most of us ūüėČ ]. You come to a new school or university and for sure, every single one of them uses a different learning platform like¬†Moodle,¬†OpenOLAT¬†or Blackboard. These applications¬†offer different ways for managing classes/courses, knowledge, vary greatly in features. They do have a few infamously¬†commonalities, such as lacking simplicity, focus and consistency. However, in this particular post I would like to take a look at Grand Valley State Universities’s Blackboard (myBB) solution. As part of the¬†CIS 623 class I am taking I received an assignment to design an improved version of it.

Why are online learning

platforms so frustrating?

In order to work on a new design I tried to locate a few of the major problem areas. There are several platform-external and -internal reasons why they might fail from my very personal opinion, such as:

Read More

First design steps… – Retirement Calculator

In order to broaden my skills I decided to take a graphical user interface class at GVSU¬†this fall. Our first assignment was to create a¬†retirement¬†calculator.¬†It was required to deliver a solution that is¬†both simple enough for someone who is completely new to these type¬†of calculations, but offers a lot of functionality to ¬ępro¬Ľ-users.

We were asked to follow as closely as possible the way ballpark suggests these kind of computation and we were asked to implement our design. While I have a rough prototype РI would like to show you my current design and would be happy to hear your feedback.

Retirement Calculator CIS 623