When yоu reference аn аrticle, yоu dо NOT need the following:
Which оf the fоllоwing would а nurse suggest аs а lean meat?
Which оf the fоllоwing is not аn аdvаntage breastfeeding gives to the mother?
Cоlоr Visiоn Deficiency (CVD) (аlso known аs Color Blindness) is а condition at which a person is unable to differentiate between certain colors. This condition can be diagnosed using a test that is conducted on children around 5 years old. A research lab would like to develop a software for detecting CVD in children at the age of one to two years old. This is a classification problem. Each data point in the dataset represents an infant or a patient. Each patient is represented by a group of attributes and a class label. The class label can be “Yes” or “No” indicating the existence of CVD. The features used to represent each data point are determined based on the current state of knowledge about the causes of CVD. This knowledge can be summarized in these points: Mutations in the OPN1LW, OPN1MW, and OPN1SW genes are among common causes of CVD. CVD is a hereditary condition that can be passed from a parent to their offspring. Color blindness is more prevalent in males than females. According to one study (Birch 1993), 8% of males and 0.4% of females have CVD. Caucasian men tend to have higher rates of CVD. Accidents that cause brain or eye injuries might lead to the development of CVD. Certain medications can cause nervous system damage and might lead to developing CVD. Using this information, a dataset is built and it looks as follows: Patient ID OPN1LW mutation OPN1MW mutation OPN1SW mutation Father has CVD Mother has CVD Sex Race Accidents Medications CVD 1 2 3 4 (P.S.: This table just shows how the dataset table would look like with its rows and columns. You do NOT need to fill in the table values.) This is the description of the different attributes: Patient ID: a unique ID number for each patient OPN1LW, OPN1MW, and OPN1SW mutations: each of these three attributes take a value of +ve or –ve, indicating whether the patient has a mutation in the corresponding gene. Father/Mother has CVD: each of these two attributes take a value of Yes or No, indicating whether the corresponding parent has CVD. Sex: Can be male or female Race: represents the race of the patient, such as Hispanic, white, black, etc. Not all patients provide this information. Accidents: represents the number of accidents that the patient had and could have caused some brain or eye injury. Medication: represents the medications that the patient took such as antibiotics or blood pressure regulators. There might be some missing values for this attribute as patients might not provide the complete list of medications. CVD: the class label, can be “Yes” or “No”. Q1. Specify the type of each of the features/attributes listed above. (8 points) Q2. Considering Decision Trees and KNN classifiers, explain the challenges and strengths of each classifier when applied to this problem. Specifically, explain how well each classifier can handle these three factors: Noise, redundant attributes, and Missing Values. The Medication and Accidents attributes contain noise. How do the different classifiers handle noise? (7 points) The three attributes representing gene mutations may represent redundancy in cases of acquired CVD in which the condition is developed due to environmental causes. (7 points) Think about the ability of the classifiers to deal with missing attribute values. Can a Decision Tree classify an instance that has some missing attribute values? For KNN, can you compute the similarity between two data points when some attribute values are missing? Can you think of a way by which each method can handle missing values? (7 points)