For this reason, if you are one of those people who advocate putting an end to Europe as a project and a socioeconomic reality together, I will simply tell you not to doubt that, if we stop being a superpower, it is very likely that we will end up as a mere area of ​​influence of another superpower. Do not judge me personally for making this statement: History is the best demonstration that, throughout the centuries, this is the reality that ends up imposing itself in the face of a leadership vacuum. If you prefer that this leadership be replaced by that of a United States that now looks more only towards itself, by that of Russia, or even by that of distant China, it is something that I leave to your subjective reflection.

So I say goodbye to you today, doubting whether in the not too distant future I may already write to you from these pages in Newspeak. The level of resistance is already set by you and your consciences, or rather, what remains of them.

“1984” is one of the essential novels of fantastic literature that I have not read. I recognize my resistance to self-flagellation, so I was not attracted to it.

But I also believe that Europe, its future rather, will not go down that path. In this sense, I am in favor of other types of outings that other authors have ventured. The common characteristics are usually (I make a Europe-Earth relationship, since obviously in this type of literature planets and not regions are treated), having reduced its extension to the minimum expression, but also the maintenance of its independence, a balancing position between the rest of the powers, a generator of culture and a crucial point of trade between all existing actors. A certain sometimes almost atavistic respect on the part of the rest of the powers is also common.

I think that view is closer to probability. Not because Europe does not decline, grow old and stop advancing. Rather because no matter how bad it is, it continues and will continue to be the icing on the cake, the first prize and whoever owns it will have many ballots to take control, which would always make the rest of the powers react by allying themselves to prevent Europe from falling into other hands. At the same time, the European political tradition, its experience in achieving balances, would facilitate this characteristic.

But this is not the only disagreement I have with the approach of the article, that although some may find it less original than it appears, at least I particularly find the point of view of how it is treated.

The reasons for this secular subsistence are the following2:  Only the printed page accompanied by illustrations is truly effective when the information is extremely complex.  The use of the book does not depend on any auxiliary or reproductive equipment.  Reading as a process of assimilation of information occurs faster than listening. However, among the most important inconveniences related to the transmission of knowledge through books, it is possible to cite the long delay between the preparation of the original and its publication. From the formal point of view, books3 are documents with more than 48 pages that constitute a bibliographic unit. By their content, books have a scientific, practical or literary character; they cover all the cultural manifestations of man. The book is the publication of a text or graphic representations greater than 48 (the cover and title page are excluded). It can consist of one volume or several, numbered or edited as part of a collection or series. According to Mikhailov, Chiornii and Guilariesvskii4, the generalizations of scientific knowledge that characterize books give them great importance, since they facilitate the investigation of the cardinal problems of any field of knowledge and constitute an incomparable means of education, teaching and culture. Books belong to the set of monographic publications and can also be included within traditional documents. In this context, monographic publications3 are defined as those that contain text, illustrations or both, in a directly readable form and are presented in full in a single volume or must be completed in a finite number of volumes (Reference adapted from ISBD (M)).

In general, they are called books. This class of documents also includes brochures, textbooks, manuals, monographs, theses, reports and reports whose definitions appear in Annex 1. The Latin American and Caribbean Information System in Health Sciences establishes a division of the documents different for the materials processed by LILACS5, however, it does not contradict in any way with that of this manual. Such categorization is the following: monographs, theses, dissertations, unconventional documents and conferences. It is necessary to clarify that although unconventional theses, dissertations and documents conform to the definition of monographs, they are part of a subset of them: in this case, manuscripts. Regarding reports and memories, their membership in periodical publications is established if their frequency is fixed.

The statistic behind this task is the Gaussian distribution, or normal distribution. The standard deviation is another of the necessary statistical concepts.

Predictive analysis builds models to infer the response of a sample (dependent variable) given a series of variables that define it (independent variables). In most cases we use artificial intelligence algorithms, which learn from historical data for which we do know the answer. Some of the most popular algorithms in this context are: decision trees, random forests, support vector machines (SVM), neural networks, boosting or linear regression. They are supervised algorithms because if we determine the answer, unlike clustering and profiling.

Predictive analysis can infer about the response to a certain campaign, estimating the predisposition of an individual to respond positively or not to a campaign. The classification or scoring algorithms determine to which specific class a given individual belongs, and the probability with which it can belong to the class.

The recommendation of friends, also known as link prediction, is another predictive analysis task that uses existing connections between the individuals that represent the sample in order to suggest new connections.

Regression is a classic case of a technique used within the field of predictive analysis. One of the most documented cases within the ecommerce scene is that of Target, which was a pioneer in identifying pregnant users according to their browsing pattern and thus recommended products especially for this audience.

The resulting model is the regression function that establishes an exit point given some input coordinates. Geometrically it is a line equidistant from all the individuals that represent the population.

There are some statistical concepts related to regression: r-squared (r ^ 2), f-test, and t-test. The r-square serves to know to what extent the line adapts to the population distribution (fitness). The f-test determines the statistical significance, establishing that the result does not happen by chance (also used to conclude an AB or MVT test). The t-test is used to know which are the variables that define the population are really relevant and which are the redundant ones (also useful for data filtering tasks -data-reduction) -.

Another of the classic tasks is the recommendation of products based on the analysis of the shopping basket, also called market-basket analysis or co-occurrence grouping. This technique is a progression of the previous one, focused only on observing which are those products that usually appear together in a transaction.

This is the case of the Amazon book recommender who tells us “Users who have bought book 1 have also bought book 2”, or buyers who have bought soda R have also bought snack S.

To solve this task, associative rules are used as a form of information representation. From the construction of these rules, the support, confidence and lift parameters are obtained. The support represents the number of times that certain items appear together in the transaction dataset. The confidence determines the frequency with which a given item appears another. While the lift establishes the probability with which a given product can appear another.

So far the best way I have found to implement this kind of analysis is by using Rattle.

Rattle is a graphical interface that is installed on top of the R statistics software. The good thing about Rattle is that it raises its interface by following the steps that the entire data mining process has, from data processing to the construction of predictive models, also including basic or descriptive statistics functions.

She joined DBi in 2009. She has worked with numerous clients since then. Among them, the UOC, Vueling or Gas Natural.

Save my name, email, and website in this browser for the next time I comment.

I mean big data.

I think big data is the best defense against big brother, first foundation speakers, or the thought police. I am based on two premises:

The first in that big data tries to obtain information using databases that do not have the purpose of storing information directly related to the object of study. That is, it seeks to obtain patterns or conclusions based on information not related in an obvious way, so publishing information may mean making it possible to know something totally unrelated a priori. Therefore, the only possible information control would be the absence of any type of information, even false information will result in true relationships.

The second is that the origin of the information does not matter, the important thing is that the more data the better, the more sources the better, the more aspects of society the better. Precisely for its objective of finding new relationships, the more it is intended to control, the more data is provided.

An example of how useless and wrong information ultimately provides true information is the LinkedIn job posting system. Currently there is talk of new professions. Articles are published on what people will be doing in 5 years and it is announced that many of these professions do not exist today. All of this is true. That could encourage people to orient their training towards specializing in these new professions (among which is big data itself). But if we combine this information with job offers, including those of LinkedIn itself, it can be seen that there is currently no such demand for new professions.

It is possible that these days have led to the accumulation of news that allows us to think that the big brother is more real than we imagine, but everything has its inertia and changes of course may be delayed but at the same time they can be very difficult to redirect. In 1956 the EEC Treaty was signed, then came the Single Act, the Maastrich Treaty, the single currency, which has been fiercely defended by all countries during this crisis despite its failures. I am optimistic and I believe that the foundations have been laid so that, once we have emerged from the crisis, the next milestone in European construction, social Europe and European taxation will be addressed.