GamesNewsCity : News and Headlines from the Video Games Industry

Machine Learning Overview
Machine learning is a branch of artificial intelligence that has applications in the finance, agricultural, educational and health industries. The jobs that require it as a requirement include software engineer, software developer and computational linguist.

Definition
 Machine learning is a branch of artificial intelligence that deals with the study and making of systems that builds models from inputted data.
It involves the use of algorithms and mathematical models to help train data that improves the output of the systems.
The phrase "machine learning" was first coined by Arthur Samuel in 1952.

Types of Machine Learning Techniques
 Supervised Learning: Algorithms are trained using input where the desired output is already assigned i.e. labeled examples. Examples include identification of faces from images, programming of autonomous cars, email spam classification or temperature prediction.
 Unsupervised Learning: This involves the use of algorithms to operate inputs where the desired output is assigned i.e. unlabeled examples. Examples include grouping of documents with related topics, news aggregator websites.
 Reinforcement Learning: This is an intermediate between supervised and unsupervised learning. It involves the receipt of feedback only after the output for a specified input or dataset has been selected. It is based on the Markov decision processes and it is applied in games whereby there is win or lose outcomes.

Applications of Machine Learning in Different Industries
 Finance industry: Machine learning helps financial companies in the areas of algorithmic trading, loan underwriting, portfolio management, fraud detection, hedge fund news analysis, and customer services.

 Agricultural industry: Machine learning is of importance when it comes to precision agriculture. It is also useful to farmers for disease detection, crop yield prediction, recognition of appropriate herbicides for different weeds, soil conditioning and analysis of chemical components of crop samples.
 Educational industry: Machine learning has been used to enhance crowd-sourced tutoring and provide virtual assistance to students. It has also saved teachers' time and made learning easier.
 Health industry: Machine learning has been used to improve cancer diagnosis and outcomes. It has also been applied to process medical images at high speeds. Machine learning has been incorporated into robotic surgery and generation of dataset towards individualized management of different diseases.

Specific Jobs in which It's a Requirement
 Software engineer: Knowledge of machine learning will enable individuals to write codes that enhance the development of algorithms on a deep level.
 Software developer: Knowledge of machine learning enables these individuals to create flow charts that makes coding for software possible.
 Data scientist: Knowledge of machine learning enables the data scientist to make informed decisions when programming languages with statistics are to be used for analysis of large dataset volumes.
 Computational linguist: Knowledge of machine learning is required for these individuals in order to effectively create rules that would help computers reproduce learnt patterns of speech.

Macro Trends in E-Commerce
Macro trends in e-commerce include the way a few companies such as Amazon and Alibaba dominate cross-border shopping, and the consistent, global growth of mobile shopping.

Macro Trend 1: Cross-Border E-Commerce Is Increasing, and Is Dominated by a Few Key Companies
In 2016, Amazon, eBay and Alibaba accounted for 65% of all cross-border purchases. That figure stayed consistent at two-thirds in 2017. Since then, these companies have continued to expand their reach into new countries. This sort of consolidation is useful when navigating the complexities of customs.
Last year, Alibaba took that competition up a notch, by expanding to retailers outside of China, while Amazon closed in China.
Shoppers are also increasingly looking for online products from outside their country. This is significant because such shoppers tend to be further along the digital or online learning process. Cross-border shopping goes beyond usual online shopping behavior and depends more on confidence in a positive delivery experience.
In 2016, 57% of respondents in a survey that included 26 countries, said they had bought from an overseas retailer in the last six months.
In the US, shoppers only buying domestically, online, decreased from 69% in 2016 to 66% in 2018.
In Europe, purchases from sellers in other European countries increased from 29 % in 2014 to 35 % in 2019, while purchases from outside the EU increased from 17 % in 2014 to 27 % in 2019.
India is one of the countries or regions with the highest rate of overseas purchase rate, at 74% of survey respondents. Italy topped that, with 79% of respondents having purchased from overseas online retailers in the past 6 months.
The outcome of this macro trend for consumers is an increased access to products that in the past were unavailable, and greater choices. Further, in developing countries, middle and upper class consumers are turning to overseas markets for high quality goods and a greater assortment of them. Consumers in wealthier countries also have access to goods from foreign companies, often at much lower prices than they would pay locally. Both sets of consumers have to consider currency exchange rates and how favorable they are. People in Denmark, for example, take advantage of lower overseas tax rates.
Retailers, on the other hand, to adapt to this trend, have to learn about overseas markets and their delivery infrastructure, financial systems, and customs regulations. They also have to consider positive after-sales experiences for refunds or exchanges.
The long term growth of cross-border e-commerce can be supported by a consolidation of delivery services that match the consolidation of retail companies.
For the IPC, there is still significant growth potential in the cross-border e-commerce market. It is dependent on reducing cross-border barriers, and is at around a quarter to a third of consumers. That is, those consumers would consider such purchases in the future.
 
Source Macro Trend 2: Mobile Shopping Is Constantly Growing

• Worldwide sales made by mobile have increased by 15% since 2016. Global sales were predicted to reach $3 trillion this year.
• Last year, the prediction was that by 2021, mobile sales would make up 73% of e-commerce sales.
• Currently in the United States a majority of customers shop using mobile devices and mobile sales have doubled since 2015. In Europe, 55% of customers are shopping using mobiles.
• This growth is driven by an increased trust in online shopping, the increased integration of cell phone usage into everyday lives, and the increased purchasing power of millennials and generation Z, who are comfortable with such technology.
• The impact on businesses of this trend is that they need to ensure they have mobile sales available, website mobile optimization, and that such features are continuously improved.
• Companies also need to be mindful of App Store Optimization — techniques to make sure their apps are more visible in app store search results.
• Businesses are also having to ensure they have various mobile payment options, such as e-wallets.
• The impact of mobile shopping is linked to increased omnichannel shopping, with customers often looking up more information. This means that mobile commerce has the potential to increase upselling and cross-selling. Mobile shopping is also having an impact on the increased demand for chat bots and it begs the need for more centralized loyalty programs.

What are macro trends for digital fraud?
The rise in digital transactions has opened several doors for hackers to commit fraud. Digital fraud macro trends are increasing and include areas involving digital or card-not-present (CNP) transactions as well as account takeover (ATO) fraud.


Global Card-Not-Present (CNP) transaction fraud is increasing

• As of 2018, Juniper Research has predicted that CNP fraud will result in substantial losses of up to $71 billion on a global scale. China and North America make up 80 percent of these fraud cases.
• In many developed countries, CNP transactions account for 60 to 70 percent of all card fraud cases. In the UK, e-commerce spending makes up approximately £248 billion and in 2016, fraud losses reached £309 million.
• Fraud losses associated with CNP transactions in the US have reached more than 50 percent of total cases of fraud loss.
• CNP transactions are expected to grow according to the US Federal Reserve Payments Study. Remote credit card payments increased by 1.5 percent from 2015 to 2016. For this same period, debit card payments and general purpose card payments increased by 1.4 percent and 3.1 percent, respectively.
• The chances of fraud have been shown to be higher with higher value remote transactions. Global transactions of $500 or more had a 22-times higher fraud rate compared to transactions of $100 or less.
• In Australia, CNP fraud resulted in approximately A$443 million in fraud loss. Out of all card fraud incidents, 80 percent of them were associated with CNP fraud between July 2016 and June 2017.
• According to the South African Banking Risk Information Centre, credit card fraud increased by 1 percent in South Africa. CNP fraud is the leading type of digital fraud, which accounts for up to 72.9 percent of credit card-related losses.
• Digital money transfer losses are forecast to grow by 130% from 2020 to 2024 due to vulnerabilities involving SIM swapping and stolen identities.
• Research reports recommend the use of an omnichannel fraud approach to lessen the impact of CNP fraud. Enhanced cybersecurity at access points and advanced machine learning may be important aspects of CNP fraud prevention.

Account takeover (ATO) fraud is accelerating

• According to the DataVisor Fraud Index Report, ATO fraud accounted for $4 billion in losses for 2019. The eCommerce sector has been targeted with ATO fraud making up 40 percent of total losses in this area.
• Fraudulent use of account numbers increased from 39.2 percent to 44.2 percent from 2015 to 2016 and is expected to grow at a faster rate.
• Account takeovers usually occur as a result of data breaches. Data breach-related ATO fraud has accounted for up to $3.3 billion in global losses.
• ATO fraud is also affecting mobile usage. Between 2018 and 2019, mobile ATO fraud increased by 45%, accounting for 679,000 incidents, according to a Javelin study.
• Businesses and consumers are heavily impacted by ATO fraud with victims having to pay $263 out-of-pocket to fix an ATO. Not only are impacts seen on a financial basis, but they’re also warranting extra time and stress to handle the problem.
• Fraudulent activity in ATO situations is not often detected until much later. It may take an average of 53 days to detect ATO fraudulent activity compared to an average of 30 days for other types of fraud.
• Banks are starting to utilize enhanced security measures with 28 percent of them prioritizing security improvements and 77 percent implementing AI solutions.
• According to a Grand View Research report, companies are integrating authentication solutions such as voice biometrics and multifactor authentication to detect fraud. The authentication solutions segment has a market size of $5 billion as of 2018.

Digital Identification Trends
Digital identification is used to gain access to healthcare, finance, and education, among many other essential services. Trends surrounding digital identification include the use of distributed ledger technologies and the use of artificial intelligence and machine learning.

Use of Distributed Ledger Technologies

•  Distributed ledger technologies are increasingly being used in digital identification. An example of these technologies is blockchain, which is used to ensure secure digital identity.
• Digital identification requires various sets of sensitive information, raising concerns regarding data breaches as well as cyberattacks.
• For instance, India’s Aadhaar, the world’s biggest national digital ID system, has suffered serious security breaches in the past that resulted in a wide exposure of confidential information.
• Blockchain permits decentralized identities where users develop their identities, register authenticating elements, and have a trusted third party to verify their particulars before they are stored in the blockchain.
• This trend is motivated by the need to decentralize the database of user identities as a strategy to reduce serious security risks if the information gets compromised.
• This is considered a trend because its recent adoption permits federated identity management and hence it is increasing the security of digital identification systems.
• Authenteq, an Iceland-based ID verification platform, is an example of a digital identity platform that is at the forefront of this trend.
•  IBM is an example of a company that has adopted blockchain technology in its digital identification management.

Use of Artificial Intelligence and Machine Learning
Artificial intelligence and machine learning are the latest tools used by both defenders and hackers in digital identification.
According to the World Economic Forum, artificial intelligence is currently the weapon most exploited by cybercriminals.
For instance, in March 2019, cybercriminals successfully used AI-based software to convincingly mimic the voice of the CEO of a UK-based energy firm. They asked for a fraudulent transfer of €220,000.
With AI-enabled capabilities, malicious people can bypass cybersecurity controls and impersonate targeted victims.
Artificial intelligence is considered a trend because, together with new encryption technologies, it is changing the landscape of digital identification for both digital attackers and defenders.
Businesses, particularly in the financial sector, are using artificial intelligence and machine learning to get smarter about fraud in digital identification.
Glenn Larson, an expert in digital imaging, indicates that Al and its machine learning subsets are enabling accurate processing, verification, and authentication of digital identities at scale.
The use of artificial intelligence ensures that digital data is converted into machine learning algorithms and then turned to accurate models. Unauthorized access or wrong login information is detected as false identification and returned as an error.
Al and machine learning in digital identification are helping in the establishment of a more robust system whose operation is devoid of reliance on human verification experts. The impact of this is faster and more secure verification of digital identity.

Digital Identification Companies
Some of the digital identification companies include ID-Pal, Ping Identity, Trulioo, Yoti, Gemalto, LifeLock, Veridium, Payfone, ForgeRock, and HooYu.

ID-Pal
ID-Pal is a digital identity verification software that is used by businesses to verify the identity of their clients in a simple, secure and convenient way.
The software was developed in 2016. The software is available for download on the Apple store and Google play store.
The company website can be accessed here.

Ping Identity
Ping Identity provides identity management software for companies and government organizations.
The company was founded by Andre Durand and Bryan Field-Elliot in 2002.
Ping identity is recognized as one of the global pioneers of identity security solutions.
The company website can be accessed here.

Trulioo
Trulioo is an online identity verification service provider. The company developed its identity verification software to help businesses comply with anti-money laundering.
Currently, the company offers its services to businesses to help them meet their "compliance, risk mitigation, and age verification needs."
Trulioo was founded by Stephen Ufford and Tanis Jorge in 2011.
Trulioo company website can be accessed here.

Yoti
Yoti was developed by Duncan Francis, Noel Hayden, and Robin Tombs. The app was launched in 2014.
Yoti is a digital identity verification software that provides a safe and convenient way to prove the identity of individuals online and in person.
The software is used by businesses to "verify identity online, improve the online safety of their users, and face to face identity checks."
The company website can be accessed here.

Gemalto
Gemalto is one of the globally recognized digital identity companies. The company operates as a digital security service provider that offers data identity and data protection services.
Gemalto was founded in 1979. The company's identity security services are used by businesses and the government to "authenticate identities and protect data. "
The company website can be accessed here.

LifeLock
LifeLock was founded by Robert Maynard and Todd Davis in 2005.
The company is recognized as an industry leader in identity security services. The company developed an identity protection app that is used by companies and individuals to protect "identity, devices and online privacy."
The company website can be accessed here.

Veridium
Veridium is a privately owned company that was founded by Hector Hoyos in 2013.
The company is a provider of identity and access management services with a focus on biometrics.
Veridium software provides employee verification, customer verification, and transaction verification solutions to businesses.
The company website can be accessed here.

Payfone
Payfone software was developed by Mike Brody and Rodger Desai in 2008.
The company is globally recognized as a digital identity authentication service provider.
The company offers mobile and digital identity authentication solutions to businesses. These services help to "validate identities, build trust and drive higher customer satisfaction."
The company website can be accessed here.

ForgeRock
ForgeRock offers digital identity management solutions to companies.
The company solutions help businesses to interact with customers and employees in a secure environment.
ForgeRock operates as a privately-owned company. The company was founded in 2010 and is headquartered in San Francisco, US.
The company website can be accessed here.

HooYu
HooYu is a global identity verification service provider. The software uses "digital footprints, ID document authentication, and facial biometrics to prove customer identity."
The software is used by retail companies, banks, and fintech to maximize customer on-boarding and increase know your customer (KYC) compliance. It is also used by fraud investigators to detect fraudulent activities.
The company website can be accessed here