I encountered some problems in registering a kaggle account. The following are the specific problems and solutions. I hope that friends who encounter the same problem will not step on the pit again ~
1kaggle mailbox cannot verify You did not enter the correct captcha response. Please try again
At first I thought it was a mailbox problem, but after changing the mailbox, the problem was still not solved
I found it on the Internet, there are two main solutions (1) Modify the hosts file (2) Install the "Google Access Assistant" plugin
Practice has proved that the second method is easier to use.
2 Speedml error when installing python
Because I am a novice, I first read the tutorials related to the Titanic: Machine Learning from Disaster project
Then I want to see how others have solved this problem, so I went to this website https://github.com/Speedml/notebooks/tree/master/titanic
Related programs need to call Speedml, a third-party library of Python.
Okay, then install it. Note that this third-party library depends on other libraries, as follows
First, it is automatically installed under Pycharm and an error is reported;
Then manually install it again and report an error;
I think it may be that the version of python installed on my computer is too messy (there are 2.7, 3.5, and there is anaconda and direct python). So I uninstalled pycharm, anaconda re-downloaded and installed python3.6
Then installed pip with cmd
Installed the above libraries in turn
Among them, installing xgboost is more troublesome, you can download the source code and install it by referring to http://blog.csdn.net/sb19931201/article/details/52236020
But this method of downloading and compiling source files with git was not successful
Finally, download the whl file to install
Download address https://download.lfd.uci.edu/pythonlibs/ghckc96n/xgboost-0.6+20171121-cp36-cp36m-win_amd64.whl
Everything goes well ~
After installing the dependent library, still use pip method to install Speedml
Titanic: Machine Learning from Disaster
Titanic: Machine Learning from Disaster
