Intelligent System FP – 08/01/2020 update
In last week, we already made a function that generates every possible moves including each move’s heuristic values for the AI, by the time this eight post published, we have finished the mini-max algorithm and improved the possibilities generator speed by using alpha-beta pruning.
Get the whole story
Intelligent System FP – 19/12/2019 update
In the seventh week of our work in doing the Intelligent System final project, we made the move or turn generator for the AI. The AI will use this generator function to generate every possible move for the player and the opponent.
The generate function will produce sets of possible movement in a form of that is similar to a tree. The function will also generate the heuristic values of a generated board using the heuristic function made in the last two weeks. Unfortunately, because this function works exponentially, the deeper the root of the tree the slower the algorithm works.
Intelligent System FP – 12/12/2019 update
In this sixth post, we implemented how the AI in our Connect 4 game will generate their own heuristic value of each column in the board, in which the AI will use to determine which column is the best to insert the disk.
We made the function of the AI’s turn from the copy of the player’s turn function. The difference is that the AI will generate heuristic values on its own simulation board. The heuristic values will be returned from the functions made in last week.
As mentioned in the previous post, the last disk inserted is used to determine the heuristic value of each column. In this case, the AI will simulates on its own game board then tries inserting the disk to calculate the heuristic value for each column. The AI will then compare every column to decide which one has the best value.
Intelligent System FP – 05/12/2019 update
In the previous post, we explained that the game has no winner jut yet. To remedy this, we implemented a function that checks other indices for similar disks. The function works by determining whether the last inserted disk and the other indices surrounding it results in a four in a row of similar disks, which then the game will be won by the disks owner.
The function also has several copies of itself that works the same way but different is that it checks two in a row and three in a row, these function is used later to determine the heuristic value of the last placed disk in a column.
Intelligent System FP – 28/11/2019 update
In the forth week, we managed to make the board to be filled with each player’s disk. In this case, there are player one and player two, in which the disks will replace the zeroes. The function will keep checking whether if the index contains a zero or other value, if it is not then that row is filled and the new disk will be placed above it instead. If the column is full the player will be noticed and asked to pick another column. We also made the main function using a simple variable to make the players change turn to fill the slots on the board. The game could not determine if there are disks that is already four in a row, so there is no winner yet.
Intelligent System FP – 21/11/2019 update
At the third week meeting, our group had taken the first step in creating the Connect 4 game. We arranged how the data and its container would be placed and presented. The data will be inserted into an array that simulates columns and rows as in the game. For this reason, the array is in the form of a nested list; the outer array simulates the columns, while the inner array simulates the board rows. The array will initially be filled with zeros, zeroes indicate that a slot is still empty and can be filled by color disks belonging to either player one or player two.
Intelligent System FP – 31/10/2019 update
The second discussion group has taken place. For our AI Connect Four, we will use the Mini-Max algorithm. Because, the algorithm is suitable for decision making to find the optimal move in a board game such as Connect Four. However, Mini-Max can be considered slow because at each turn, players have many choices. Thus, many branch factors must be sought to find the optimal move.
Later after that, we also got a new idea for our final project. Because we agreed that one of the techniques of machine learning, clustering, is quite useful and is widely applied in everyday life. It is possible for us to replace our project into a program that manages daily schedules automatically with the help of machine learning, which we call AI planning.
Intelligent System FP – 24/10/2019 update
This is the first update for the Intelligent System Final Project.
Group members: Arkaan, Farras, Fiqhy
In our first group meeting, we discussed what project that needs a machine learning capability and suitable to be the final project for Intelligent System course.
The term AI is widely applied in games, therefore, our group decided to make a computer AI for a game. The game we chose was Connect Four, because we thought this project could be considered simple yet difficult and challenging enough to work on.
Intelligent System – Week 5 (UNFINISHED)
In this post, is written my understanding of unsupervised learning through observation: Clustering. The clustering technique in this post will use K-means and Apriori algorithms.
Also, both the exercise and assignments of week five are uploaded here.
Assignment : Question -> Exercise for K-Means Answer -> K-MEANS(assignment)
Exercise : K-MEANS -> K-MEANS(exercise) Apriori -> Exercise for Apriori
What is DNS Spoofing?
DNS spoofing is a part of computer hacking in which searched domain names are diverted to some other incorrect IP address due to which the traffic of the victim’s system is diverted to attacker’s system. Using DNS spoofing poison is injected into the address resolution protocol of the victim.
Consider that an attacker started DNS spoofing on the system of a victim and diverted the IP of facebook.com to the attacker’s own IP address. If the victim tries to open facebook.com, the DNS will open the attacker’s IP instead of facebook.com. If the attacker does that, there is a big probability of stealing data or cookies from the victim’s system.
Typograph is a standards compliant theme with a JQuery powered tabbed sidebar box and an ad under the first post on the index page. This theme has no images and is purely based on CSS elements and typography. Ideal for future customization. Typograph was styled "from the ground up" on a highly customized version of the Sandbox theme.
Designed by Morten Rand-Hendriksen - designer, information philosopher and author based out of Burnaby, BC.
You can change the contents of the tabbed box by editing the tabbedBox.php file located in the Typograph theme directory.
