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 traffic intensity in networking, Explain Traffic Intensity in netwo...

Explain Traffic Intensity in networking. Traffic Intensity: The traffic load on a given network may be on interoffice trunk lines, the local switching unit or other common s

Coupling and cohesion, Coupling and cohesion can be shown using a:- Dep...

Coupling and cohesion can be shown using a:- Dependence matrix

What is called when network address prefixed by 1110, Network address prefi...

Network address prefixed by 1110 is a? Network address prefixed through 1110 is a multicast address.

Applications of parallel processing, Applications Of Parallel Processing ...

Applications Of Parallel Processing Parallel computing is an development of serial computing that effort to emulate what has always been the affirm of affairs in the natural wo

Combinational logic, At a particular moment in time the output of a COMBINA...

At a particular moment in time the output of a COMBINATIONAL logic circuit depends uponthe inputs to the circuit at the same instant.(Unlike sequential logic circuits where the out

Observations of high level language program, Q. Observations of High Level ...

Q. Observations of High Level Language Program? Observations Integer constants appeared nearly as frequently as structures or arrays. Most of the scalars were foun

Explain instruction level of parallel processing, Instruction Level It ...

Instruction Level It refers to condition where different instructions of a program are implemented by different processing elements. Most processors have various execution unit

Prototyping and incremental development, a) Write  the main differences amo...

a) Write  the main differences among prototyping and incremental development.    b) Explain the commonality and main differences among agile approach and RUP.

Examples of input, (i)  Input a single ASCII character into BL register wit...

(i)  Input a single ASCII character into BL register without echo on screen  CODE SEGMENT  MOV AH, 08H;         Function 08H  INT 21H          ;         the character inpu

Minimis logic function using K-map, Minimise the logic function F (A, B, C,...

Minimise the logic function F (A, B, C, D) = Π M (1, 2, 3, 8, 9, 10, 11,14) ⋅ d (7, 15) Use Karnaugh map.   Ans. Given function F=∏M(1,2,3,8,9,10,11,14).d(7, 15) F'=B'D+B'C+AC+AB'

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