- Yujiao li added an answer:2Could synthetic control method be optimized in another way?
- Marion G Ceruti added an answer:4Given a new user in a network, how can we find his/her actual publications in DBs?
- Dennis Weyland added an answer:2How does the Harmony search work for feature selection?
- Adel Sabry Eesa added an answer:5Could any one tell me whether ordinal attribute can be given as an input to SVM and MLP?
- Preeti Balaji added an answer:10Does anyone have experience with SAR image classification?
- Shaveta Chutani added an answer:11Can anyone help with features of an image?
- Priyanka Shah added an answer:4Can molecules from Pubchem or ChEMBL can be used as decoy in classification studies?
- Tannaz Akbarpour added an answer:10Is there any multi-class SVM classifier available in MATLAB?
- Patricia Ryser-Welch added an answer:4Does anybody know what the main approaches of reinforcement learning in continuous state and action spaces are?
- Angel Martinez-Tenor added an answer:8What is the difference between value iteration and policy iteration methods in reinforcement learning?
- Xin Ye added an answer:3Anyone know about machine learning tool for urdu treebank?
- Mostafa Ali Shahin added an answer:4Any tool for Speech to Text conversion especially for Indian English?
- Avinash Dudi added an answer:4I have two sets of data, healthy and damaged; what can I do with this?
- Amir Jalilifard added an answer:8How can I study the past spending behaviour of a customer in a banking perspective and predict the next purchase category and amount of buy?
- Vladimir Batagelj added an answer:1How do I normalize values of structural distance-based measure?
- Aasim Khurshid added an answer:4What is the running time complexity of SVM and ANN?
- Said Jadid Abdulkadir added an answer:11Which are the new trends in the development of ANNs?
- Sachin Patil added an answer:18When and why do we need data normalization?
- Antonio Miguel added an answer:6What is the difference between mutual information and information gain?
- Ransalu Senanayake added an answer:7How can I cluster users based on their posts/queries?
- Miguel Hernandez-Silveira added an answer:3How can I evaluate the performance (weights and bias) given from a nonlinear SVM (Support Vector Machine) Kernel Model?
- José Francisco Moreira Pessanha added an answer:4What are your suggestions for Clustering Binary Categorical data?
- Haroldo Fraga de Campos Velho added an answer:8In choosing an activation function for SVR or Neural network, what is the advantage/disadvantage of radial base function to using a sigmond function?
- Jorge Lopez-Cifre added an answer:1Anyone familiar with ICLR papers? (They are all available at arxiv.org)
- İzzet Pembeci added an answer:8How should I proceed to implement NLP (almost) from scratch (collobert 2011)?
- Katharina Morik added an answer:15What is the best algorithm for supervised text categorization of papers?
- Mohammad Mahdi Momenzadeh added an answer:7Is there a high-resolution training image database for pattern recognition purpose?

I am learning the "synthetic control method(SCM)" since I am doing comparative case study. I have questions about its reliability in optimization.

After ranking the similarity of every control unit with treatment unit, I choose different top N sample in this Sythetic Control (SCM). When I iterate the sample size from 2-280 in this method(Synth package in R), the objective function(MSE) is fluctuated very much. This unstable result make SCM unreliable. After placebo test, I delete the outliers, MSE got smaller and stable,but the optimized weight of control group and predictors are still fluctuate and non-converge as sample size increase.

My question is whether this non-convergence is because the initial value in the optimization due to genetic algorithm? Or because of potential collinear between predictors?

I have two solutions for your suggestions:

(1)I applied another similarity measure matrix --mahalanobis distance instead of Euclidean distance which consider the correlation between predictors.

(2) Considering the potential non linear relationship, I used neural-network, Support vector machine, Gradient boosting regressiong tree for optimize the weight. But the first 2 methods turn out the overfitting and only GBRT works fine. Could I use this result and move forward? I am not sure this is the right track.