Machine Learning : Cross Validation Technique

-
Cross validation is one way to avoid over fitting of data in machine learning .
Cross validation works in such way :

1st Way : One Round Cross Validation :  in this data is divided into  50-50 portion ,one portion is for test set , one portion is validation of model that is prepared with the help of 50% of data .

Leave one out cross validation(LOOCV)  :

 In this ,we leave one data set for cross validation  and rest data set for use for training of model .

k-fold cross validation : In this k data set are used to trained  the model ,while k-1 are using for validate or test  the model .


Comments

Post a Comment

Popular posts from this blog

Install Mysql from ZIP without MySQL Installer or Exe

-
Image
How to install MySQL from Zip file . Step1 : Download  MysSQL   from  myql  website ..    Download From Here you will see  mentioned  folder  and  file .except red circle folder ,that I have  created manually  . Data: for storing database file . error: for storing error  log . tmp: it need to  for installation  of myql Step  2:  now create  a configuration file  my.cnf(linux)  ,my.ini (windows) . Mandatory configuration parameter that need to include in your file as shown . Step 3 : Now  Need to initialize  mysqld  by mentioned any of command .         mysqld    --initialize         mysqld    --initialize-insecure In my case first command did not  work properly  ,but 2 nd command work file ..  ,now  I am telling difference between these command . , first command will generate a  secure key for connecting mysql  but 2 nd command will let you intilize mysql  without password , you can generate it letter  my mentioned command Alter user  ‘root’@’localhos
Read more

Logging in Python .

-
Python logging is as much important for python as flavours in Ice Cream , lies in politics and life without goal .without flavour ice cream means ,it may be tasty for you but not for others ,so  a programme should be like that it should be understandable by others .Logging is used for log purpose of a programme ,what actually it is doing .lies in politics means you can not grow your code without logging otherwise ,it is very difficult to track your programme . So before further moving we can see .where logging can be used . 1: Displaying a information of  programme . 2: Issuing a warning . 3: Catching  a error . 4: Catching Critical information . How we can  achieve by python logging: So for achieving we need to know  ,level in python ,mostly 5 type of level are used in python DEBUG,INFO,WARNING ,ERROR and CRITICAL .let's elaborate what actually each level means in python . DEBUG : Debug we used for diagnostic purpose or detailed information . INFO     : Programme is w
Read more

Dictionary in Python

-
# how to create blank dictionary dictionary = {} print(dictionary) # declare dictionary with key and value dictionary = {"fname": "avni","mname":"singh","lname":"rathor"} print(dictionary) # accessing key and value from dictionary for key in dictionary:     print(key)     print(dictionary[key]) # accessing key and value with items method for key,value in dictionary.items():     print(key,value) #  value can also be fetched from get method print(dictionary["fname"]) print(dictionary.get("fname")) # now will add key and value in dictionary . dictionary["age"] = "31" print(dictionary) # now removing item fname  from dictionary dictionary.pop("fname") print(dictionary) # del keyword can also be used del del dictionary["lname"] print(dictionary) # can clear dictionary with clear function dictionary.clear() print(dictionary) # dict constructior
Read more