matlab - Classifier with a vector of features and a matrix of classes -

i'm new in classification i'm asking advice on how start.

i've created matlab script create 2 matrices, 1 class identifier, meaning 100x1 contains group data is. group 1 (1) or group 2 (2).

the second matrix contains features 100x40 40 features each point.

what's best way start, i'm lost. matlab has functions can use?

i appreciate help.

thank you.

it depends on version of matlab using, best starting point @ statistics toolbox supervised learning. here starting tips matlab 2013a:

http://www.mathworks.co.uk/help/stats/supervised-learning.html

let's assume data is

classes: 100x1 features: 100x40 

for each method, first line shows how fit classification model , second lines shows how classify first row of data in features.

statistics toolbox

naive bayes classification

wikipedia: https://en.wikipedia.org/wiki/naive_bayes_classifier

myclassifier = naivebayes.fit(features, classes) myclassifier.predict(features(1,:)) 

nearest neighbors

wikipedia: https://en.wikipedia.org/wiki/nearest_neighbour_classifiers

myclassifier = classificationknn.fit(features, classes) myclassifier.predict(features(1,:)) 

classification trees

wikipedia: https://en.wikipedia.org/wiki/classification_tree

myclassifier = classificationtree.fit(features, classes) myclassifier.predict(features(1,:)) 

support vector machines

wikipedia: https://en.wikipedia.org/wiki/support_vector_machine

note support vector machines moved 2013a bioinformatics toolbox , supports classification 2 groups.

myclassifier = svmtrain(features, classes) svmclassify(myclassifier, features(1,:)) 

discriminant analysis

wikipedia: https://en.wikipedia.org/wiki/discriminant_analysis

myclassifier = classificationdiscriminant.fit(features, classes) myclassifier.predict(features(1,:)) 

neural network toolbox:

if have 2 classes, use neural network toolbox pattern recognition typing nnstart