Cloud Hosting

Tensorflow icon

Machine learning has many practical applications that you can use in your projects or on the job.

In the Machine Learning with Python Certification, you'll use the TensorFlow framework to build several neural networks and explore more advanced techniques like natural language processing and reinforcement learning.

You'll also dive into neural networks, and learn the principles behind how deep, recurrent, and convolutional neural networks work.

TensorFlow is an open source framework that makes machine learning and neural networking easier to use.

The following video course was created by Tim Ruscica, also known as Tech With Tim. It will help you to understand TensorFlow and some of its powerful capabilities.

Not PassedNot Passed0/32

Neural networks are at the core of what we call artificial intelligence today. But historically they've been hard to understand. Especially for beginners in the machine learning field.

Even if you are completely new to neural networks, these video courses by Brandon Rohrer will get you comfortable with the concepts and the math behind them.

Not PassedNot Passed0/4

Machine learning has many practical applications. By completing these free and challenging coding projects, you will demonstrate that you have a good foundational knowledge of machine learning, and qualify for your Machine Learning with Python certification.

See the rest here:
Machine Learning with Python Certification | freeCodeCamp.org

Solving novel problems and setting a new milestone in competitive programming.

Creating solutions to unforeseen problems is second nature in human intelligence a result of critical thinking informed by experience. The machine learning community has made tremendous progress in generating and understanding textual data, but advances in problem solving remain limited to relatively simple maths and programming problems, or else retrieving and copying existing solutions. As part of DeepMinds mission to solve intelligence, we created a system called AlphaCode that writes computer programs at a competitive level. AlphaCode achieved an estimated rank within the top 54% of participants in programming competitions by solving new problems that require a combination of critical thinking, logic, algorithms, coding, and natural language understanding.

In our preprint, we detail AlphaCode, which uses transformer-based language models to generate code at an unprecedented scale, and then smartly filters to a small set of promising programs.

We validated our performance using competitions hosted on Codeforces, a popular platform which hosts regular competitions that attract tens of thousands of participants from around the world who come to test their coding skills. We selected for evaluation 10 recent contests, each newer than our training data. AlphaCode placed at about the level of the median competitor, marking the first time an AI code generation system has reached a competitive level of performance in programming competitions.

To help others build on our results, were releasing our dataset of competitive programming problems and solutions on GitHub, including extensive tests to ensure the programs that pass these tests are correct a critical feature current datasets lack. We hope this benchmark will lead to further innovations in problem solving and code generation.

Continue reading here:
Competitive programming with AlphaCode - DeepMind

Even as the worlds leading militaries race to adopt artificial intelligence in anticipation of future great power war, security forces in one of the worlds most conflict-prone regions are opting for a more measured approach. In Africa, AI is gradually making its way into technologies such as advanced surveillance systems and combat drones, which are being deployed to fight organized crime, extremist groups, and violent insurgencies. Though the long-term potential for AI to impact military operations in Africa is undeniable, AIs impact on organized violence has so far been limited. These limits reflect both the novelty and constraints of existing AI-enabled technology.

Artificial intelligence and armed conflict in Africa

Artificial intelligence (AI), at its most basic, leverages computing power to simulate the behavior of humans that requires intelligence. Artificial intelligence is not a military technology like a gun or a tank. It is rather, as the University of Pennsylvanias Mark Horowitz argues, a general-purpose technology with a multitude of applications, like the internal combustion engine, electricity, or the internet. And as AI applications proliferate to military uses, it threatens to change the nature of warfare. According to the ICRC, AI and machine-learning systems could have profound implications for the role of humans in armed conflict, especially in relation to: increasing autonomy of weapon systems and other unmanned systems; new forms of cyber and information warfare; and, more broadly, the nature of decision-making.

In at least two respects, AI is already affecting the dynamics of armed conflict and violence in Africa. First, AI-driven surveillance and smart policing platforms are being used to respond to attacks by violent extremist groups and organized criminal networks. Second, the development of AI-powered drones is beginning to influence combat operations and battlefield tactics.

AI is perhaps most widely used in Africa in areas with high levels of violence to increase the capabilities and coordination of law enforcement and domestic security services. For instance, fourteen African countries deploy AI-driven surveillance and smart-policing platforms, which typically rely on deep neural networks for image classification and a range of machine learning models for predictive analytics. In Nairobi, Chinese tech giant Huawei has helped build an advanced surveillance system, and in Johannesburg automated license plate readers have enabled authorities to track violent, organized criminals with suspected ties to the Islamic State. Although such systems have significant limitations (more on this below), they are proliferating across Africa.

AI-driven systems are also being deployed to fight organized crime. At Liwonde National Park in Malawi, park rangers use EarthRanger software, developed by the late Microsoft co-founder, Paul Allen, to combat poaching using artificial intelligence and predictive analytics. The software detects patterns in poaching that the rangers might overlook, such as upticks in poaching during holidays and government paydays. A small, motion-activated poacher cam relies on an algorithm to distinguish between humans and animals and has contributed to at least one arrest. Its not difficult to imagine how such a system might be repurposed for counterinsurgency or armed conflict, with AI-enabled surveillance and monitoring systems deployed to detect and deter armed insurgents.

In addition to the growing use of AI within surveillance systems across Africa, AI has also been integrated into weapon systems. Most prominently, lethal autonomous weapons systems use real-time sensor data coupled with AI and machine learning algorithms to select and engage targets without further intervention by a human operator. Depending on how that definition is interpreted, the first use of a lethal autonomous weapon system in combat may have taken place on African soil in March 2020. That month, logistics units belonging to the armed forces of the Libyan warlord Khalifa Haftar came under attack by Turkish-made STM Kargu-2 drones as they fled Tripoli. According to a United Nations report, the Kargu-2 represented a lethal autonomous weapons system because it had been programmed to attack targets without requiring data connectivity between the operator and munition. Although other experts have instead classified the Kargu-2 as a loitering munition, its use in combat in northern Africa nonetheless points to a future where AI-enabled weapons are increasingly deployed in armed conflicts in the region.

Indeed, despite global calls for a ban on similar weapons, the proliferation of systems like the Kargu-2 is likely only beginning. Relatively low costs, tactical advantages, and the emergence of multiple suppliers have led to a booming market for low-and-mid tier combat drones currently being dominated by players including Israel, China, Turkey, and South Africa. Such drones, particularly Turkeys Bakratyar TB2, have been acquired and used by well over a dozen African countries.

While the current generation of drones by and large do not have AI-driven autonomous capabilities that are publicly acknowledged, the same cannot be said for the next generation, which are even less costly, more attritable, and use AI-assisted swarming technology to make themselves harder to defend against. In February, the South Africa-based Paramount Group announced the launch of its N-RAVEN UAV system, which it bills as a family of autonomous, multi-mission aerial vehicles featuring next-generation swarm technologies. The N-RAVEN will be able to swarm in units of up to twenty and is designed for technology transfer and portable manufacture within partner countries. These features are likely to be attractive to African militaries.

AIs limits, downsides, and risks

Though AI may continue to play an increasing role in the organizational strategies, intelligence-gathering capabilities, and battlefield tactics of armed actors in Africa and elsewhere, it is important to put these contributions in a broader perspective. AI cannot address the fundamental drivers of armed conflict, particularly the complex insurgencies common in Africa. African states and militaries may overinvest in AI, neglecting its risks and externalities, as well as the ways in which AI-driven capabilities may be mitigated or exploited by armed non-state actors.

AI is unlikely to have a transformative impact on the outbreak, duration, or mitigation of armed conflict in Africa, whose incidence has doubled over the past decade. Despite claims by its makers, there is little hard evidence linking the deployment of AI-powered smart cities with decreases in violence, including in Nairobi, where crime incidents have remained virtually unchanged since 2014, when the citys AI-driven systems first went online. The same is true of poaching. During the COVID-19 pandemic, fewer tourists and struggling local economies have fueled significant increases, overwhelming any progress that has resulted from governments adopting cutting-edge technology.

This is because, in the first place, armed conflict is a human endeavor, with many factors that influence its outcomes. Even the staunchest defenders of AI-driven solutions, such as Huawei Southern Africa Public Affairs Director David Lane, admit that they cannot address the underlying causes of insecurity such as unemployment or inequality: Ultimately, preventing crime requires addressing these causes in a very local way. No AI algorithm can prevent poverty or political exclusion, disputes over land or national resources, or political leaders from making chauvinistic appeals to group identity. Likewise, the central problems with Africas militariesendemic corruption, human rights abuses, loyalties to specific leaders and groups rather than institutions and citizens, and a proclivity for ill-timed seizures of powerare not problems that artificial intelligence alone can solve.

In the second place, the aspects of armed conflict that AI seems most likely to disruptremote intelligence-gathering capabilities and air powerare technologies that enable armies to keep enemies at arms-length and win in conventional, pitched battles. AIs utility in fighting insurgencies, in which non-state armed actors conduct guerilla attacks and seek to blend in and draw support from the population, is more questionable. To win in insurgencies requires a sustained on the ground presence to maintain order and govern contested territory. States cannot hope to prevail in such conflicts by relying on technology that effectively removes them from the fight.

Finally, the use of AI to fight modern armed conflict remains at a nascent stage. To date, the prevailing available evidence has documented how state actors are adopting AI to fight conflict, and not how armed non-state actors are responding. Nevertheless, states will not be alone in seeking to leverage autonomous weapons. Former African service members speculate that it is only a matter of time before before the deployment of swarms or clusters of offensive drones by non-state actors in Africa, given their accessibility, low costs, and existing use in surveillance and smuggling. Rights activists have raised the alarm about the potential for small, cheap, swarming slaughterbots, that use freely available AI and facial recognition systems to commit mass acts of terror. This particular scenario is controversial, but according to American Universitys Audrey Kurth Cronin, it is both technologically feasible and consistent with classic patterns of diffusion.

The AI armed conflict evolution

These downsides and risks suggest the continued diffusion of AI is unlikely to result in the revolutionary changes to armed conflict suggested by some of its more ardent proponents and backers. Rather, modern AI is perhaps best viewed as continuing and perhaps accelerating long-standing technological trends that have enhanced sensing capabilities and digitized and automated the operations and tactics of armed actors everywhere.

For all its complexity, AI is first and foremost a digital technology, its impact dependent on and difficult to disentangle from a technical triad of data, algorithms, and computing power. The impact of AI-powered surveillance platforms, from the EarthRanger software used at Liwonde to Huawei-supplied smart policing platforms, isnt just a result of machine-learning algorithms that enable human-like reasoning capabilities, but also on the ability to store, collect, process collate and manage vast quantities of data. Likewise, as pointed out by analysts such as Kelsey Atherton, the Kargu 2 used in Libya can be classified as an autonomous loitering munition such as Israels Harpy drone. The main difference between the Kargu 2 and the Harpy, which was first manufactured in 1989, is where the former uses AI-driven image recognition, the latter uses electro-optical sensors to detect and hone in on enemy radar emissions.

The diffusion of AI across Africa, like the broader diffusion of digital technology, is likely to be diverse and uneven. Africa remains the worlds least digitized region. Internet penetration rates are low and likely to remain so in many of the most conflict-prone countries. In Somalia, South Sudan, Ethiopia, the Democratic Republic of Congo, and much of the Lake Chad Basin, internet penetration is below 20%. AI is unlikely to have much of an impact on conflict in regions where citizens leave little in the way of a digital footprint, and non-state armed groups control territory beyond the easy reach of the state.

Taken together, these developments suggest that AI will cause a steady evolution in armed conflict in Africa and elsewhere, rather than revolutionize it. Digitization and the widespread adoption of autonomous weapons platforms may extend the eyes and lengthen the fists of state armies. Non-state actors will adopt these technologies themselves and come up with clever ways to exploit or negate them. Artificial intelligence will be used in combination with equally influential, but less flashy inventions such as the AK-47, the nonstandard tactical vehicle, and the IED to enable new tactics that take advantage or exploit trends towards better sensing capabilities and increased mobility.

Incrementally and in concert with other emerging technologies, AI is transforming the tools and tactics of warfare. Nevertheless, experience from Africa suggests that humans will remain the main actors in the drama of modern armed conflict.

Nathaniel Allen is an assistant professor with the Africa Center for Strategic Studies at National Defense University and a Council on Foreign Relations term member. Marian Ify Okpali is a researcher on cyber policy and the executive assistant to the dean at the Africa Center for Strategic Studies at National Defense University. The opinions expressed in this article are those of the authors.

Microsoft provides financial support to the Brookings Institution, a nonprofit organization devoted to rigorous, independent, in-depth public policy research.

Continued here:
Artificial Intelligence Creeps on to the African Battlefield - Brookings Institution

Machine learning has been applied to many health-related tasks, such as the development of new medical treatments, the management of patient data and records, and the treatment of chronic diseases. To achieve success in those SOTA applications, we must rely on the time-consuming technique of model building evaluation. To alleviate this load, Yue Zhao et al have proposed a PyHealth, a Python-based toolbox. As the name implies, this toolbox contains a variety of ML models and architecture algorithms for working with medical data. In this article, we will go through this model to understand its working and application. Below are the major points that we are going to discuss in this article.

Lets first discuss the use case of machine learning in the healthcare industry.

Machine learning is being used in a variety of healthcare settings, from case management of common chronic conditions to leveraging patient health data in conjunction with environmental factors such as pollution exposure and weather.

Machine learning technology can assist healthcare practitioners in developing accurate medication treatments tailored to individual features by crunching enormous amounts of data. The following are some examples of applications that can be addressed in this segment:

The ability to swiftly and properly diagnose diseases is one of the most critical aspects of a successful healthcare organization. In high-need areas like cancer diagnosis and therapy, where hundreds of drugs are now in clinical trials, scientists and computationalists are entering the mix. One method combines cognitive computing with genetic tumour sequencing, while another makes use of machine learning to provide diagnosis and treatment in a range of fields, including oncology.

Medical imaging, and its ability to provide a complete picture of an illness, is another important aspect in diagnosing an illness. Deep learning is becoming more accessible as data sources become more diverse, and it may be used in the diagnostic process, therefore it is becoming increasingly important. Although these machine learning applications are frequently correct, they have some limitations in that they cannot explain how they came to their conclusions.

ML has the potential to identify new medications with significant economic benefits for pharmaceutical companies, hospitals, and patients. Some of the worlds largest technology companies, like IBM and Google, have developed ML systems to help patients find new treatment options. Precision medicine is a significant phrase in this area since it entails understanding mechanisms underlying complex disorders and developing alternative therapeutic pathways.

Because of the high-risk nature of surgeries, we will always need human assistance, but machine learning has proved extremely helpful in the robotic surgery sector. The da Vinci robot, which allows surgeons to operate robotic arms in order to do surgery with great detail and in confined areas, is one of the most popular breakthroughs in the profession.

These hands are generally more accurate and steady than human hands. There are additional instruments that employ computer vision and machine learning to determine the distances between various body parts so that surgery can be performed properly.

Health data is typically noisy, complicated, and heterogeneous, resulting in a diverse set of healthcare modelling issues. For instance, health risk prediction is based on sequential patient data, disease diagnosis based on medical images, and risk detection based on continuous physiological signals.

Electroencephalogram (EEG) or electrocardiogram (ECG), for example, and multimodal clinical notes (e.g., text and images). Despite their importance in healthcare research and clinical decision making, the complexity and variability of health data and tasks need the long-overdue development of a specialized ML system for benchmarking predictive health models.

PyHealth is made up of three modules: data preprocessing, predictive modelling, and assessment. Both computer scientists and healthcare data scientists are PyHealths target consumers. They can run complicated machine learning processes on healthcare datasets in less than 10 lines of code using PyHealth.

The data preprocessing module converts complicated healthcare datasets such as longitudinal electronic health records, medical pictures, continuous signals (e.g., electrocardiograms), and clinical notes into machine learning-friendly formats.

The predictive modelling module offers over 30 machine learning models, including known ensemble trees and deep neural network-based approaches, using a uniform yet flexible API geared for both researchers and practitioners.

The evaluation module includes a number of evaluation methodologies (for example, cross-validation and train-validation-test split) as well as prediction model metrics.

There are five distinct advantages to using PyHealth. For starters, it contains more than 30 cutting-edge predictive health algorithms, including both traditional techniques like XGBoost and more recent deep learning architectures like autoencoders, convolutional based, and adversarial based models.

Second, PyHealth has a broad scope and includes models for a variety of data types, including sequence, image, physiological signal, and unstructured text data. Third, for clarity and ease of use, PyHealth includes a unified API, detailed documentation, and interactive examples for all algorithmscomplex deep learning models can be implemented in less than ten lines of code.

Fourth, unit testing with cross-platform, continuous integration, code coverage, and code maintainability checks are performed on most models in PyHealth. Finally, for efficiency and scalability, parallelization is enabled in select modules (data preprocessing), as well as fast GPU computation for deep learning models via PyTorch.

PyHealth is a Python 3 application that uses NumPy, scipy, scikit-learn, and PyTorch. As shown in the diagram below, PyHealth consists of three major modules: First is the data preprocessing module can validate and convert user input into a format that learning models can understand;

Second is the predictive modelling module is made up of a collection of models organized by input data type into sequences, images, EEG, and text. For each data type, a set of dedicated learning models has been implemented, and the third is the evaluation module can automatically infer the task type, such as multi-classification, and conduct a comprehensive evaluation by task type.

Most learning models share the same interface and are inspired by the scikit-API learn to design and general deep learning design: I fit learns the weights and saves the necessary statistics from the train and validation data; load model chooses the model with the best validation accuracy, and inference predicts the incoming test data.

For quick data and model exploration, the framework includes a library of helper and utility functions (check parameter, label check, and partition estimators). For example, a label check can check the data label and infer the task type, such as binary classification or multi-classification, automatically.

PyHealth for model building

Now below well discuss how we can leverage the API of this framework. First, we need to install the package by using pip.

! pip install pyhealth

Next, we can load the data from the repository itself. For that, we need to clone the repository. After cloning the repository inside the datasets folder there is a variety of datasets like sequenced based, image-based, etc. We are using the mimic dataset and it is in the zip form we need to unzip it. Below is the snippet clone repository, and unzip the data.

The unzipped file is saved in the current working directory with the name of the folder as a mimic. Next to use this dataset we need to load the sequence data generator function which serves as functionality to prepare the dataset for experimentation.

Now we have loaded the dataset. Now we can do further modelling as below.

Here is the fitment result.

Through this article, we have discussed how machine learning can be used in the healthcare industry by observing the various applications. As this domain is being quite vast and N number application, we have discussed a Python-based toolbox that is designed to build a predictive modelling approach by using various deep learning techniques such as LSTM, GRU for sequence data, and CNN for image-based data.

Read the original:
How to build healthcare predictive models using PyHealth? - Analytics India Magazine