imgaboy
Member
- Joined
- Feb 7, 2018
- Member Type
- Student or Learner
- Native Language
- Hungarian
- Home Country
- Serbia
- Current Location
- Serbia
Hi there!
I do a research and I have finished my work, but my mentoree wants me to make some correction in my work.
I have finished four paragraph. Has anybody enough time to help me with a grammar check.
Thanx
Here is it:
To deeply understanding of the data, we had to carry out further preliminary tasks, which revealed properties hidden up to this point. As the Fig. 1. presents the noisy and complex nature of this set of data, made impossible to use simple statistical or clustering methods to create a predictive model. To deeply understand the user behaviour and attitudes, we had made other cleaning processes and transformation on our data.
To investigate the structure of the data and understand the user behaviour, we had to find a visualization which solves that problem. In this step, the biggest gap was the unbalanced nature of the data (n “Failed” = 419, n “Completed” = 184) which overcomplicated the interpretation of the diagrams. To solve that situation we attempt to show our results by using diagrams which showing density distribution.
This paper described a statistical methodology for predicting binary outcome in a set of data which was created in short Mooc and driven by the teacher. Based on these data, we found a successful model which was influenced by a couple of strong features. As we could see on the fig. 6 in the first 30 most weighted features contain enormous data about time, mouse click, mouse move and distance.
Features
We defined 263 features to describe our clickstream data. There was a two type of data. In the first group, there was the data which was collected during the filling process in the incoming test. The second type was the clickstream which was collected during the learning process in the three parts of the curriculum (see Table 1.). These collection we putted into 18 main categories: binary and numeral answers provided to input and output tests (28+60), time spent on the quizzes (6) and the sites where was the curriculum (7), number of visits on the site of quizzes (6) and site of curriculum (7), mouse move distance in pixels (13), average mouse speed (13), cumulated data (6), number of mouse movement on a page (13), number of clicks on a page (13), use of test buttons during the input/output testing (4), number of scrolls on a page (13), the last login date to a page compared to the first login to the site (13), the first login date compared to the first login to the site (13), days spent on the sites (13), mean behaviour on the sites (19), number of calendar days passed between the output tests (7), binary output results (1) , output results (2), user related data (6).
I do a research and I have finished my work, but my mentoree wants me to make some correction in my work.
I have finished four paragraph. Has anybody enough time to help me with a grammar check.
Thanx
Here is it:
To deeply understanding of the data, we had to carry out further preliminary tasks, which revealed properties hidden up to this point. As the Fig. 1. presents the noisy and complex nature of this set of data, made impossible to use simple statistical or clustering methods to create a predictive model. To deeply understand the user behaviour and attitudes, we had made other cleaning processes and transformation on our data.
To investigate the structure of the data and understand the user behaviour, we had to find a visualization which solves that problem. In this step, the biggest gap was the unbalanced nature of the data (n “Failed” = 419, n “Completed” = 184) which overcomplicated the interpretation of the diagrams. To solve that situation we attempt to show our results by using diagrams which showing density distribution.
This paper described a statistical methodology for predicting binary outcome in a set of data which was created in short Mooc and driven by the teacher. Based on these data, we found a successful model which was influenced by a couple of strong features. As we could see on the fig. 6 in the first 30 most weighted features contain enormous data about time, mouse click, mouse move and distance.
Features
We defined 263 features to describe our clickstream data. There was a two type of data. In the first group, there was the data which was collected during the filling process in the incoming test. The second type was the clickstream which was collected during the learning process in the three parts of the curriculum (see Table 1.). These collection we putted into 18 main categories: binary and numeral answers provided to input and output tests (28+60), time spent on the quizzes (6) and the sites where was the curriculum (7), number of visits on the site of quizzes (6) and site of curriculum (7), mouse move distance in pixels (13), average mouse speed (13), cumulated data (6), number of mouse movement on a page (13), number of clicks on a page (13), use of test buttons during the input/output testing (4), number of scrolls on a page (13), the last login date to a page compared to the first login to the site (13), the first login date compared to the first login to the site (13), days spent on the sites (13), mean behaviour on the sites (19), number of calendar days passed between the output tests (7), binary output results (1) , output results (2), user related data (6).
