forestry industry, as well as agriculture. this system started working in the fall of twenty-three. the decree was signed by vladimir putin. exporters will sell proceeds at least until may 2025. japan has introduced new anti-russian sanctions. this time , alrosa, tupolev, mikron and 39 other companies were subject to restrictions. the blacklist also included enterprises from friendly countries. these are kazakhstan, india, china, uae and uzbekistan. all of them are according to the japanese version... said the head of the bank, german gref, as part of annual meeting of shareholders. according to him, the latest changes in the tax system will make their own adjustments, but the bank is ready for this. at the end of 23 years, the bank contributed more than 900 billion to the budget. this year the figure will be

higher, gref emphasized. let me remind you that in 5 months sber earned almost 630 billion rubles in net profit, an increase over the year of more than 6.5%. and washington prohibits the use of kaspersky lab antiviruses in the united states; its software poses risks to the country’s national security. this was reported on the website of the us department of commerce. any transactions with the russian company will be stopped in july; in september all software will be banned. spersky, the organization itself intends to take measures to maintain contacts with american partners; in other countries, the company’s business remains stable. it was economic news, briefly.

the ministry of emergency situations is mastering neural networks. russian rescuers are now monitoring natural fires using high technology. the intelligent system was created at skoltech and is already being used in practice. about this within the framework of the st. petersburg international economic conference. form

said deputy prime minister of russia dmitry chernyshenko. neuron automatically predicts the situation 5 days in advance throughout russia with an accuracy of more than 84%. this solution is already being used by the ministry of emergency situations in the irkutsk region and krasnoyarsk territory, helping to save forests, money and lives. natural fires are an eternal problem of the summer season in russia. every year, the fire element costs the country's budget billions of rubles. people are dying, houses are burning down, and air quality is deteriorating. it is not surprising that the most advanced technology available to humanity, artificial intelligence, but not only fire threatens russians, catastrophic downpours, squalls, heat waves, cold snaps, all these are the realities of the new climate era. is it possible to predict these phenomena using neural networks? how does a model predicting wildfires work, and can it warn about other natural disasters with the same accuracy?

in general, in many artificial intelligence machine learning tasks , the actual data needs to be collected and prepared, and what data is this? firstly, this is data from thermal points, that is, where there was an excess of temperature, in what area of ​​space, at what point in time, this is satellite data, secondly, this is data that is also obtained from satellite data on the underlying surface, various navigation indices, of course, this terrain data, of course this is distance data. to populated areas,

because in general in predicting fires, you need to understand that the human factor, unfortunately, sometimes happens and does take place, and finally, it’s different. two, three ahead in a certain area of ​​space, it is potentially high, yeah, this is what, so to speak, is fed to this model as an input, that array of data that is analyzed at each specific moment in time to make a forecast, but the model needs train, set up these knobs, and for this there should be, so to speak, markings, that

is, this is historical information about those fires where they actually occurred, having already collected this array of data, in our case it’s 10 years, we all regions of russia such array. 100 options, well, of course, there are all sorts of models like the nesteru index, the soil moisture index, which, naturally, are used in the structures of those involved, this is a very simple combination, literally a couple of factors, but if this is higher, and this is lower, then it means most likely there

could potentially be problems here, but in reality there is a much larger situation in practice where something can happen, but non-russia is able to get this information from this array of historical data and then use similar ones, so to speak, with large errors it is still established from satellite data, but if you aggregate it into your model, it’s not, well, i understand that it is one of the main parameters from it, yes, you probably have a very important parameter, so if it bad according to satellite data initially, this is how much it harms you,

or or the other parameters outweigh it turns out quite well, well, fortunately, we have access to the model, we add data that - firstly, represents a characteristic of the weather here now, historical data about the weather there at previous points in time, because this model is fed with data for the last conditionally 7 days, yes, because we need to track the dynamics, right? and secondly, that is, this is data about the weather, and secondly, this is data about the weather forecast, which is considered to be a calculation model in a special way. and you somehow calculate the weather forecast in your own way, or take the results, we use the modeling results of roshydromet in this case, unfortunately.

firstly, it meets their needs, that is, they can already use it are used to carry out some of their daily tasks, but if , of course, we had weather data and weather forecasts with greater detail, then we would even achieve much higher accuracy, well , yes, if i understand correctly, there in general, the scale, you probably need a spatial resolution of literally kilometers for this to be good, because there...

on the scale of russia, ranking the forces of means , after all, we are now talking about such a global one and this is a slightly different story in this sense there are several tens of a couple of tens of kilometers there - as they say already. this is this type of model, one that predicts a fire 5 days in advance,

that is, every day in the corresponding grid cells that covers the entire territory of russia, we predict the corresponding probability, this is allowed, this is this model, it is embedded in the internal control ministry of emergency situations, passed these acceptance tests there, now they use it on a daily basis.

some files, raives, uh, some other government agencies, here’s how this will affect them, on

work, it is still premature to say, because the model was literally recently introduced and passed the corresponding stages of acceptance of operation. here are other natural disasters, that is, your model can be adapted to predict something else, but i think, well, we really have intense precipitation, but here i think it’s up to the meteorologist.

also mathematical models based on all sorts of differential equations, they in some sense may be less accurate, because they are not adapted to specific the data that is observed here, now on a specific river, is to a lesser extent, at least, adapted, but they internally include, as it were, physical equations, well , how our world as a whole is structured, due to this combination of such models and models, which are trained on this mireset will allow you to obtain much higher accuracy. forecast, that is, here some additional work will be required specifically with

algorithms, a combination of models based on physics based on non-networks, we are like that we are developing technologies at the skaltihi artificial intelligence center, but that’s great, this is just such a reasonable hybrid, yes a reasonable hybrid, the second comment here is, the next step that could be taken here, in fact there is a very large field for fundamental namely research, which can then turn into something important specifically for... uh industry, namely the construction of what is now called a fundamental model of artificial intelligence, well, everyone has seen these cats who draw uh models like kandinsky and others, yes, yes, this is such a... not to say that it’s a bomb, but it looks cool, well, these are these types of technologies, they can and should be adapted for the data that we are now discussing, and from this can be very useful, moreover, abroad, in the usa, in china, this work is actively being carried out, we need to start it here too, well, yes, of course, in russia there is a real

boom in erosetey, right now other models are being developed for natural probability assessments. another one, by the way, also from skoltech, deals with long-term assessments flood risks. more details about it in our review. we load climate model calculations, weather station readings, reanalysis data into the machine, add information about the relief and voila. artificial intelligence provides an answer to the question of where, with what probability , rivers may overflow their banks in the foreseeable future. this cannot be fully called a forecast. roset will not be able to predict the fact of a flood in a particular area, but the creators of the system, skoltech specialists, say yes. in its current form , the results of artificial intelligence will be useful to financial companies and large businesses. the model will be more accurate, its metrics will be higher if we model, for example, a year in advance, by month. this format seems optimal from the point of view of the quality of the models. the first story is who this could be for, where it could be used, this is definitely the

insurance industry. colleagues can use our models to clarify the risk of natural disasters and accordingly. this will be a useful tool for them, the second industry for which the use of our models is relevant is this is banking, that is, we can reduce investment risks, including credit risks. the third industry for which our models may be relevant is business as such, that is, for example, for such industries as agriculture, mining industries, the technology is applicable and other emergency situations. in addition to floods, artificial intelligence can calculate droughts, heat waves, increased winds, convective processes, that is, heavy precipitation, as well as the degradation of the long-term world zloty, for which a separate model has been developed, however no matter what natural disasters the neural network evaluates, it is always necessary to make adjustments for climate changes, due to which the data entered into the system must be constantly adjusted; one of the levels

of input data to our model is climate scenarios, which means they are formed on... greenhouse gas emissions and are developed accordingly by climatologists ; in our case, we always take an ensemble of climate scenarios, which we use within these models and we carry out a kind of validation of them and adjustment, that is, we take a fairly large number of models and adjust them to data from the weather station, unfortunately, today this is the only way that allows us to achieve this. any precision in our modeling. the creators of the model say that at the moment it cannot replace full-fledged monitoring systems; there is not enough data to form accurate forecasts. sometimes, in order for the machine to be able to calculate at least something, you have to resort to tricks, including modeling virtual gauging stations on rivers, if it doesn’t have enough real ones. the wider the network of meteorological and hydrological observations, the sooner it will be possible to integrate

artificial intelligence into forecasting specific, rather than hypothetical, emergency situations. does not stand still, our climate is changing quite rapidly, we will not, of course, be alarmists, we will not say that this is a disaster, but nevertheless , this is how much, we just talked about all sorts of twist handles in models, how much really this problem is acute when you generally, in principle, we we assess any risks, it doesn’t matter.

are used to mitigate them for some other general actions to do, which are important in these situations regarding twisting the handles. in fact, the current practice of working with non-networks is that large non-network models, they are usually the opposite, their quality only improves

as new data is added to them, in a sense, the non-network model pulls out more and more examples from them, like different combinations of features with each other interact there humidity. realizes, well, if you can say so, of course about the algorithm, that this combination corresponds to - there is no increased probability, yeah, in this sense, due to the fact that the more historical material we have, the data material on which we do all this, the more likely the potential reliability of such a system will be, well, right now... i’d like to talk about what just caused an information boom among meteorologists and climatologists, you of course heard this at the end of autumn, ah, so to speak, advanced british

minds, but that’s all it's under google was done, they created a neural network that began to make weather forecasts better than the hydrodynamic models of the european center for medium-range forecasts, and of course, with much less computing resources, it’s all cheaper, it’s all better. and of course , immediately, as happens in such cases, the thought immediately arose, some meteorologists are so decadent that that’s it, meteorologists are no longer needed, hydrodynamic models are no longer needed, that’s it, artificial intelligence, well, i’m much more careful about this, and what about you? ? i'm also very careful about these alarmist statements, a person is not going anywhere in the near future, especially a person with good expert knowledge, a person armed with such models based on artificial intelligence, can solve his problems there faster, more accurately and more problems in the same time, if we talk specifically about models

from google, or, for example, there is a company called huawei, which, based on its capabilities and developments, also made such a panga climate model, of a similar type in some sense fundamental, that means what we are talking about, first, again, data is collected, and this data is like...

if we have some initial measurements, what is the most likely trajectory, how will they develop? and computationally, already using such a model, it is much faster than if you again calculate this big climate model for the whole earth, which you do year by year, of course, one without the other still won’t go anywhere in the sense that it’s completely the right story when you're doing something rude i guess, the rough forecast is this - it can be a computationally heavy model , you calculate it, and then you refine it using... an artificial intelligence model, which is additionally adapted to a specific region, the important thing here is that it’s still expert interpretation and calculations,

an expert is needed for this, this is the first, second, here of course there is still a lot of work, i meant that in order to build such models, it is precisely the development of fundamental artificial intelligence algorithms that is required, there is a lot of work here for those who study mathematics, mathematics and algorithms, all this is based on everyone. host processes, what kolmagogorov also came up with is diffusion models, and this is where people can find themselves, namely fundamental scientists, but then this is - it’s possible not just to generate cats, which we now all see on the internet, there are very beautiful pictures, but that’s all -that’s not it, it makes our life better without cats, life is bad, but it actually wants its own industrial applications, and this is where it goes next... and it should go, even such a reasonable hybrid of artificial intelligence with hydrodynamic models, a huge number of ensemble calculations, everything, everything, all the data that exists, but they still do not

save from butterflies. error in the initial data, the response can be anything, in general, yes, probably, this is a question, a debatable question, which means the point is that in the case of neural networks, for example, you can also make an assessment of the uncertainty of the forecast , see in what situations she is very big, where smaller, this is the first one.

the planet is changing, perhaps neural networks will help us better understand these changes and adapt to them. evgeny burnaev, professor, head of the skoltech artificial intelligence center, answered science questions today. evgeniy vladimirovich, thank you very much for the interesting conversation, well, perhaps, see you again. thank you, see you again, thank you.

a massive attack was repelled without pilots... 114 drones were shot down overnight. in the krasnodar region, bapla fell on the territory of an oil refinery. in the very in krasnodar, the boiler house and bus station buildings were damaged. collective security issues are on the agenda in almaat. sergei lavrov took part in a meeting of the council of ministers of the cst country. about the main results in the live broadcast of robert frantsev. the us

is preparing to secretly deploy. nuclear weapons in britain, this is the conclusion reached by the international campaign for the elimination of nuclear weapons. such actions flagrantly violate the non-proliferation treaty. we have adapted to them; the crew can destroy it in a matter of seconds. how to protect the sky over melitopol and how russian air defense systems work. this is in the report of our correspondent from the front line. let's start with the news. kazakhstan, the conflict unleashed by the west in ukraine poses a very serious threat to the csto. the kiev regime is used as an instrument of aggression against russia. this is a statement by sergei lavrov. the head of the russian foreign ministry took part in a meeting of the foreign ministers of the csto countries in almaat, and then answered questions from journalists. and we will learn all the details from the chief of our central asia bureau, robert franz, he's in direct contact. robert, greetings.