Decision tree learning for cancer diagnosis, Computer Engineering

Assignment Help:

Assignment 1: Decision tree learning for cancer diagnosis

In this mini-project, you will implement a decision-tree algorithm and apply it to breast cancer diagnosis. For each patient, an image of a fine needle aspirate (FNA) of a breast mass was taken, and nine features in the image potentially correlated with breast cancer were extracted. Your task is to develop a decision tree algorithm, learn from data, and predict for new patients whether they have breast cancer. Dataset can be downloaded from U.C. Irvine Machine Learning Repository.

1.       Collect the data set from my website. Each patient is represented by one line, with columns separated by commas: the first one is the identifier number, the last is the class (benign or malignant), the rest are attribute values, which are integers ranging from 1 to 10. The attributes are (in case you are curious): Clump Thickness, Uniformity of Cell Size, Uniformity of Cell Shape, Marginal Adhesion, Single Epithelial Cell Size, Bare Nuclei, Bland Chromatin, Normal Nucleoli, Mitoses. (Note that the UCI document page specifies a different number of attributes, because it refers to a set of several related datasets. For detailed information of the dataset that we use here, see this document.)

2.       Implement the ID3 decision tree learner, as described in Chapter 3 of Mitchell. You may program in C/C++, Java. Your program should assume input in the above format.

3.       Implement both misclassification impurity and information gain for evaluation criterion. Also, implement split stopping using chi-square test.

4.       Divide the data set randomly between training (80%) and testing (20%) sets. Use your algorithm to train a decision tree classifier and report accuracy on test. Run the same experiment 100 times. Then calculate average test performances (accuracy, precision, recall, f-measure, g-mean).

5.       Compare performances by varying the evaluation criteria. Make a table as follows:

Evaluation Criteria

Accuracy

Precision

Recall

F-measure

G-mean

misclassification impurity

 

 

 

 

 

information gain

 

 

 

 

 

6.       Answer the following:

a.       Which evaluation criterion and confidence level work well? Why?

b.       Do you see evidence of overfitting in some experiments? Explain.

 


Related Discussions:- Decision tree learning for cancer diagnosis

Explain semaphore, What is a semaphore? Semaphore: It is a synchroniz...

What is a semaphore? Semaphore: It is a synchronization tool which gives a general-purpose solution to controlling access to critical sections.

fisherpersons are not over-fishing, Prepare the Relational Tables to signi...

Prepare the Relational Tables to signify the following situation, which is defined by means of text and an Entity Relationship Diagram. Note: Do not attempt to vary the ERD. Yo

Intuitively simple - user friendliness, Intuitively Simple - User Friendlin...

Intuitively Simple - User Friendliness This is the degree to which the system operates in congruence with human operation, in effect the machine does what we naturally think i

Selection - artificial intelligence, Selection - artificial intelligence: ...

Selection - artificial intelligence: However the first step is to choose the individuals that will have a shot at becoming the parents of the next generation. Hence this is kn

What is the maximum frequency of operation, For the circuit demonstrated be...

For the circuit demonstrated below, what is the Maximum Frequency of Operation? Are there any hold time violations for FF2? When yes, how do you modify the circuit to ignore them?

Determine the boolean expression using k-map, Reduce the following equation...

Reduce the following equation using k-map Y = BC‾D‾ + A‾BC‾D + ABC‾D + A‾BCD + ABCD Ans. Multiplying the first term with (A+A') Y = A'BC'D' + ABC'D' + A'BC'D + ABC'D + A'BCD +

What are the advantages of electronic payment system, What are the advantag...

What are the advantages of electronic payment system? Advantages of the electronic payment system are illustrated below: This gives good security schemes. Four important sec

What is dots per inch, Q. What is Dots per Inch? DPI (Dots per Inch) is...

Q. What is Dots per Inch? DPI (Dots per Inch) is a measure for actual sharpness of onscreen image. This relies on both the resolution and size of the image. Practical experienc

What is assembler, What is Assembler An assembler is a program which ta...

What is Assembler An assembler is a program which takes as input a symbolic language program and produces output as its binary machine language equivalent. The input is known a

Database, er table for hospital management system

er table for hospital management system

Write Your Message!

Captcha
Free Assignment Quote

Assured A++ Grade

Get guaranteed satisfaction & time on delivery in every assignment order you paid with us! We ensure premium quality solution document along with free turntin report!

All rights reserved! Copyrights ©2019-2020 ExpertsMind IT Educational Pvt Ltd