odizeey:
Please, I keep seeing error like this (getOption (repos)) when I want to load data set in Rstudio, what should I do please.

4kings:

getOption("repos" ) is a command in R, so i don't quite understand you.

How did you try to load your dataset and can you also post the full error shown.

J0nyb0y:
Guys can anyone help with materials on Knowledge Engineering... I think it's a branch of AI.

My project supervisor gave me as a project topic..:

Combating Security Challenges in Knowledge Engineering Products

4kings:
@odizeey you have to install the required packages.
use:
install.packages('Rcpp')

Or using devtools directly:
devtools::install_github("tidyverse/readr" )

odizeey:
thanks. I'm even thinking of using python, I want to download d anaconda file.

4kings:

Okay.
What are you doing by the way?

odizeey:
Please, I keep seeing error like this (getOption (repos)) when I want to load data set in Rstudio, what should I do please.

odizeey:
I just downloaded a dataset from kaggle, about the most complete player.
I just want to try my best, I'm just a learner. What kind of problem is that: determining the most complete player, classification, regression, or which?

Samcent:
The dataset you want to load is in which format,.... csv, xlsx...? There's a package in r called rio. Install the package if you haven't and load. Then use 'import()' function to load your dataset. Make sure the dataet is already downloaded in your machine and remember to added the right file extension. Eg: import('iris.xlsx') , for iris dataset in an excel format. I hope the above helps.

Samcent:
Can you describe the dataset? If there is a variable on the COMPLETE PLAYER that is captured as YES or NO, then it's a classification task. Logistic regression is also a classification task. If you don't mind can we work on this dataset together for learning purpose. I am also new to machine learning.

odizeey:
yes its csv, how can I install d rio package.

Samcent:
From your r studio, enter the code: install.packages('rio'). After installing, load the package with the code: library ('rio'). You can also load your csv files using 'read.csv()' function. You don't need to load any package to use read.csv function. Make sure your data is in your working directory.

odizeey:
I just downloaded a dataset from kaggle, about the most complete player.
I just want to try my best, I'm just a learner. What kind of problem is that: determining the most complete player, classification, regression, or which?

odizeey:
there is no variable indicating yes or no, just the club of the players, nation, weighg, height, and so on.
We can work on it surely. It would be nice.

4kings:

I have to know the dataset and what the task is to determine if it's classification, regression or what ever.
Classification deals with discrete data e.g Positive and Negative, while Regression deals with Continous data like predicting stock prices.

Goodluck with solution.

Samcent:
Here is a link to the dataset: www.kaggle.com/kevinmh/fifa-18-more-complete-player-dataset

4kings:

Interesting, I can see that it's more of provision of dataset and the task to perform is determined by you.

Well there were several options for exploration in this referenced project. Now depending on what you choose to explore on the dataset cluster analysis might also come in handy, but for the most part, it's a classufication problem.

You can start with algorithms you're familiar with and do a cross-validation to see which performs better and if you can you could perform some ensemble learning.

Samcent:
Well, am sorry I upset you. I was only trying to learn while helping. Thanks and have a blissful night rest.

odizeey:
I just downloaded a dataset from kaggle, about the most complete player.
I just want to try my best, I'm just a learner. What kind of problem is that: determining the most complete player, classification, regression, or which?

Samcent:
The Fifa 18 More Complete Player Dataset has 17,994 observations and 185 variables. What do you have in mind to accomplish with this dataset? Honestly, I am not sure which machine learning task to apply. However, I think doing a k-means cluster analysis may help throw some light on the data. Also, you may not need to work with all the 185 variables. These are my amateur submissions though.

Samcent:
The Fifa 18 More Complete Player Dataset has 17,994 observations and 185 variables. What do you have in mind to accomplish with this dataset? Honestly, I am not sure which machine learning task to apply. However, I think doing a k-means cluster analysis may help throw some light on the data. Also, you may not need to work with all the 185 variables. These are my amateur submissions though.

SoftEng:



I've looked through kaggle to get a feel of the Complete Player dataset. I'm guessing you're referring to one of the links down below (I'm not sure the exact one).

If the above is correct, then there is no clear cut pre-defined task for the dataset (i.e. no clearly a classification or a regression problem). Based on Exploring Data section in authors description, it may seem that they provided the dataset for mainly analysis/visualisation of the dataset. However, people can work on many different types of tasks as evident in Kernels section of each dataset link. For example, predicting player position, which is a classification task.

In case you are not sure where to start, checkout the Kernels section of each dataset. The kernel section contains codes of tasks people worked on. You can read through their code/tutorial and run the code as well. This should help you improve your skills. You can also check out the "Discussion" section of the dataset link, which should give you a some additional ideas of what people are working on, questions raised etc.


Short Note On Data Analysis/Visualisation
Data analysis/visualisation is a key preliminary step in any datascience project. It enables you (the data scientist) to get a good feel (and other useful information such as mean and variance of the dataset distribution). Additionally, through data analysis/visualisation you may be able to find out duplicate features in samples of your dataset as well as eliminate or reduce the number of features in samples of your dataset. Usually, after data analysis/visualisation stage, the data scientist is better equipped to select the right kind of machine learning model (e.g. random forest, gradient boosting, neural network and so on) that may work best for the dataset.


Links
1. https://www.kaggle.com/thec03u5/fifa-18-demo-player-dataset
2. https://www.kaggle.com/artimous/complete-fifa-2017-player-dataset-global
3. https://www.kaggle.com/kevinmh/fifa-18-more-complete-player-dataset

4kings:
Huh? I don't get, was this post meant for me?

SoftEng:
4kings and Samcent
both of you have been excellent providing help to odizeey. well done.

Samcent, I'm not sure why you posted your last comment. So far, the discussions between 4kings and yourself have been good and constructive. No cause to upset anyone. Please keep the discussions rolling.

Thank you all.

SoftEng:
4kings and Samcent
both of you have been excellent providing help to odizeey. well done.

Samcent, I'm not sure why you posted your last comment. So far, the discussions between 4kings and yourself have been good and constructive. No cause to upset anyone. Please keep the discussions rolling.

Thank you all.

raymod170:


there are options available you can apply
1. Creation of a closed domain question answering system
2. Carry out an analysis to determine which player would be highly rated in the coming year since you have access to real data from sofifa.com
3. data visualization of each country, and teams this players play for as some players in a country football also play together in the same team
4. facial recognition of players by using ml algorithm to do so
5.creation of an ultimate team using this real time data to form a team based on ratings among other things
6. Create an algorithm that looks for or creates line of dichotomy between Messi and Ronaldo based on the data available
7. a knowledge based system AI inspired

and much more

but the ideas are unlimited and this is what am talking about using the available data to create things worth it

it's a start and I believe we would get there

thanks 4kings for showing us this website as I have found some interesting datasets which I would use to create some Airtficial intelligence projects starting with this FIFA dataset

best regards

4kings:

What security challenge specifically are you dealing with?