The history of data mining

It is related to development of data mining directly that mass data accumulation became possible.
Collection of data in a digital form took that I did data analysis with a computer into consideration, and it was already performed in 1960's when I read the history of data mining.
Language SQL for relational data base and the operation appeared in 1980's, and dynamic data analysis was enabled on demand.
I reached it in 1990's, and the quantity of data increased explosively.
Data warehouse has begun to be used for accumulation of data.
With this, a general idea of data mining appears as technique to handle data in large quantities in a database, and it is the history of new data mining that technique of a statistical analysis or search technology in the field of artificial intelligence came to be applied.

A technical background of data mining

It is thought that a technical background of data mining has rapid progress and low cost of computer machinery of performance of a computer.
Environment collected data in large quantities by these phenomena was regulated well.
But it may be to be here these days that an analysis calculation came to be possible easily.
I think that words of data mining are hard to be said to still seep.
Because collection power, data processing capacity of data, data accumulation power, data analysis power are the regular present, it may be said with the technology that it came to be possible for.
The person whom even an individual started to try to do data mining appeared here and there recently.
An individual analyzes data of gambles such as finance connection such as stocks, horse racing and thinks that I will have the superiority as a thinking person appeared.

A kind of a decision tree

A decision tree has else two names.
It is not used for a recurrence tree classification, and it is used a real number value for an approximation of a function to take.
An example: An estimate of a price of a house. An estimate of hospitalization of a patient.

Classification tree y in the case of categorical variable.
For example, as a result of sex, man or woman, match, I win and lose.

Text mining

It is the general term of text data analysis technique it divides the text data which are not arranged, a normal natural sentence into words, and to get constant knowledge and idea by analyzing the appearance frequency or a correlation with technique of data mining.
google of search engine uses technology of text mining.
I worked to cut prose into pieces in every word in Japanese text mining, and it seemed that Japanese was late because morphological analysis was necessary, but it depended so that the evolution that in late years was technical and technical fluidity resembled it, and even business came to reach a practical use level to some extent.
How to use how I extract a customer and needs of a market and do it and analyze dissatisfaction point to our product into is expected by analyzing notes from a user put to a free answer, call center or a Web site of a questionnaire to qualitative analysis information, a bulletin board.
A connection of words and words understands what kind of relations occur.
In addition, I interpreted this automatically and I answered it and, for an inquiry from a customer by an E-mail, was able to be said to transfer to the person in charge concerned.

A decision tree of data mining technique

A decision tree is a graph to be decided in a field of decision theories such as risk management, and it is used I draw up a plan, and to arrive at an aim.
A decision tree is made for the purpose of helping with decision-making.
A decision tree is special form of hierarchy structure.

weka

weka is a data mining tool of an open source.
Because it is made JAVA, there is a difficulty in speed a little, but a kind of a decision tree to be able to use is having many kinds.
I have potence as it is a shame to say that I say closely with a decision tree tool.
A thing becoming a tool to look for a rule is almost prepared.
I am support multi-platforms such as Windows and Unix system, MacOSX.
It means that I preferred versatility to speed.
Central, and, as for the development, New Zealand Waikato University is performed.
I possess the tool which can appoint the tool which can carry out data mining interactively and a flow of processing by an icon connecting with drop, and it is it in the software that it is easy to use very much.
There is decision tree software elsewhere, but it is easy to read output data, too and is plain.
I use a decision tree of weka as data mining, but am held up as text mining.
Server log mining is possible, too and has various things.
But, speaking of a weak point, it is a weak point that must do optimization of data by oneself.
Software of a decision tree thinks that I am enough with one weka if I understand it what it is to do it by oneself.
Because information of weka rolls unexpectedly, it may be good to try to loiter around in a net.

An investment trust

An investment trust, investment trust and abbreviation, a lot of investors leave a fund with assets use company, and assets use company invests the fund in stocks and a bond, financial assets such as derivative financial instruments or real estate, and it is the financial instruments which distribute the profit that I got by the use between an investor.
As a matter of course, an investor will bear the loss that use caused. I call it a unit trust in the U.K. in U.S.A.

An investment trust is means of a risk hedge

There is not a capital guarantee as a general rule an investment trust invests it in financial instruments such as stocks without a capital guarantee or a bond, and to settle an account individually.
An investment trust can expect a good return than an ordinary deposit and fixed deposits such as banks, but can interpret this as receiving risk premium facing that I took a risk equivalent to.
As for the investment trust, a payoff in particular is lifted the ban on, and, under the present conditions that low interest, an interest income for a deposit by zero-interest-rate policy cannot almost anticipate, an investment trust attracts attention as means of a new risk hedge for assets use.

return and a risk of an investment trust

It was forbidden the handling of risk articles such as investment trusts in a bank and a life insurer.
Sale is not accepted, and the investment trust was a patent of a brokerage firm virtually, but a brokerage firm and an investment trust use company of a system borrow part space of a bank of an infantility from December, 1998 in what was lifted the ban on sale at a bank window by a flow of Big Bang as a start, and a type of industry that is various to a post office enters life insurance, a damage insurance company, a credit association, an end the removal of a ban in an investment trust sale window and the form that it is it now, and sale competition of an investment trust intensifies.
But it is not an investment trust, but appears in a company canceling the handling of investment trust sale like Nippon Life Insurance Co. handling an article, variable rate insurance, the strange forehead annuity insurance that resembled an investment trust in saleability.
I am multifarious by an investment of an investment trust I take how much risk, and how much return is provided.
For example, a risk has a bigger stocks than a bond, and it is assumed that a return is big.

A risk of an investment trust and a standard of a return

One of the standards that standardized a risk of an investment trust and degree of a return has the sharp ratio that Nobel prize economist Sharp developed.
I subtract an interest rate of no risk asset from a prospective return and broke this in a risk, standard deviation undertaking, and, as for the thing having a big value with plus, use will be effective.
In addition, I become gauging of a sweat shirt when I assume a denominator a beta risk.
In the case of an investment trust, an evaluation index has many the cases which a sharp ratio is used for.