House Price Prediction using Neural Networks

#1

Hi I m using neural networks to solve this problem but could not cross 64 percent accuracy.

Here is my solution.

link to the code:

https://drive.google.com/file/d/1oEQ7OuqdIb_0EowL4qqjNDuzdAXeMcE-/view?usp=sharing

#2

hey @keshari ,
i just went through your code , its pretty good . Few things that l want to tell you is

  1. Always try to optimize your code once you have completed working on it. Your code is currently very large because of outputs , so just don’t show so much of outputs .

Tips for For better performance on this dataset,

  1. Most important is feature selection and extraction , its not required that you go for imputing column values . Sometimes , if that column is not corelated to our target column ,we just drop it.

  2. Depending upon this correlation with target value we select only the useful columns/feature for our model training. and if there is any kind of thinking you get ,like if combine any two columns together we can understand the work more easily or splitting any column values , we can make it learn better , then you can surely perform this on your dataset. This thing is purely based on experience upon datasets. So think accordingly.

  3. To check , is there any requirement of scaling or normalizing data . If so , then do so.

  4. Now comes the main modelling Part.
    Depending upon dataset , we choose our algorithms .
    For current dataset ,with only 1000 rows , this is very small dataset and hence neural networks will not perform so good on it. As deep learning requires a lot of tons of data to learn more better the data.
    So you can go for other algorithms like decision trees , random forest or boosting . Various machine learning boosting techniques can help you to score very good on this. Search about lightgbm and how to use it.
    Random Forest , SVM have a shown good results on this dataset. So just try with them once.

Hope this might helps :slightly_smiling_face:.

#3

I hope I’ve cleared your doubt. I ask you to please rate your experience here
Your feedback is very important. It helps us improve our platform and hence provide you
the learning experience you deserve.

On the off chance, you still have some questions or not find the answers satisfactory, you may reopen
the doubt.