In this section, we provide the proofs to the Lemmas which were introduced in the main paper.

The ability to detect and adapt to changes in data distributions is crucial to maintain the accuracy and reliability of machine learning models. Detection is generally approached by observing the drift of model performance from a global point of view. However, drifts occurring in (fine-grained) data subgroups may go unnoticed when monitoring global drift. We take a different perspective, and introduce methods for observing drift at the finer granularity of subgroups. Relevant data subgroups are identified during training and monitored efficiently throughout the model's life. Performance drifts in any subgroup are detected, quantified and characterized so as to provide an interpretable summary of the model behavior over time. Experimental results confirm that our subgroup-level drift analysis identifies drifts that do not show at the (coarser) global dataset level. The proposed approach provides a valuable tool for monitoring model performance in dynamic real-world applications, offering insights into the evolving nature of data and ultimately contributing to more robust and adaptive models.

The Lipstick Index, as it's known, was a phrase coined by Leonard Lauder, chairman of the board at Estee Lauder, in the early noughties when it became clear in times of economic crisis, sales of color cosmetics - in particular lipstick - soar as an affordable way to treat yourself. Last year, sales of cosmetics plummeted. With bustling make-up counters in department stores closed for much of 2020, sales of designer brand cosmetics were down by more than 40% according to market research firm NPD, which equates to a loss of £500 million (around $689 million and AU$902 million). Meanwhile, sales of more affordable cosmetics in supermarkets fell by 22% in the UK, according to the Top Products survey from retail trade magazine, The Grocer, adding a further £183 million loss (around $256 million / AU$218 million). A combination of the rise of working from home and mandatory masks meant thousands of us, including this writer, ditched our make-up bags altogether.

AI, AR, VR, or any form of virtual reality, isn't a new concept within the beauty industry, but it certainly is a remarkable one. Virtual reality has always stirred a web of speculation and curiosity amongst tech-savvy enthusiasts, and whether the enriching digital-based experience was put to the test by mass brands like Nike and IKEA, or luxury fashion houses such as Gucci and Louis Vuitton, augmented reality is the once-niche concept that is turning traditional ways of shopping and experiencing products into a revolutionary trend. It isn't surprising that the pandemic has helped accelerate the digital innovation in the beauty space. With retail shops temporarily closing their doors around the world, consumers have been driven online to fuel their beauty needs. To help make the online shopping experience easier, retailers are using AI technology -- and it's working.

The next time you spot a lip color that you like, you can quickly use AI to find a corresponding shade in Chanel's range of lipsticks. The company is announcing today its new Lipscanner app that lets you use your phone's camera to identify a hue -- whether it's on someone's lips or just the color of your favorite purse. Then, it'll suggest a match from Chanel's "lip universe," which includes more than 400 products encompassing different finishes and shades. This isn't the first time a beauty company has made an app to identify lipstick shades -- YSL's Perso system will even print out the exact color you scan, making it more sophisticated. Lipscanner will also allow you to virtually "try on" the lipstick to see if you like it.

A Korean cosmetics company has taken the guesswork out of makeup. Amorepacific's Color Tailor app analyzes your face, determines what shade of lipstick would best complement your complexion and then sends the results to a machine that mixes up a custom tube of it while you wait. The company is showing off its Lip Factory system next week at the Consumer Electronic Showcase, the annual tech convention that's moved to the virtual world in light of the pandemic. Users upload a photo of their face onto the Color Tailor app, which uses artificial intelligence to select a flattering hue from among more than 2,000 different possible combinations. The machine then mixes and matches pigments to create a custom product you can buy.

Data privacy is an increasingly important aspect of many real-world big data analytics tasks. Data sources that contain sensitive information may have immense potential which could be unlocked using privacy enhancing transformations, but current methods often fail to produce convincing output. Furthermore, finding the right balance between privacy and utility is often a tricky tradeoff. In this work, we propose a novel approach for data privatization, which involves two steps: in the first step, it removes the sensitive information, and in the second step, it replaces this information with an independent random sample. Our method builds on adversarial representation learning which ensures strong privacy by training the model to fool an increasingly strong adversary. While previous methods only aim at obfuscating the sensitive information, we find that adding new random information in its place strengthens the provided privacy and provides better utility at any given level of privacy. The result is an approach that can provide stronger privatization on image data, and yet be preserving both the domain and the utility of the inputs, entirely independent of the downstream task. Increasing capacity and performance of modern machine learning models lead to increasing amounts of data required for training them (Goodfellow et al., 2016). However, collecting and using large datasets which may contain sensitive information about individuals is often impeded by increasingly strong privacy laws protecting individual rights, and the infeasibility of obtaining individual consent.

New Delhi: The pace of innovation in beauty tech is accelerating, but the innovations in beauty retail today don't fully solve the customer's requirements, believes Ritika Sharma, Founder, and CEO, House of Beauty. The beauty and wellness company recently launched a new e-commerce platform Boddess.com. IANSlife spoke to Sharma to know how Augmented reality (AR) and artificial intelligence (AI) are redefining the concept of beauty tech and if it's the future. Q: How are you using AI and AR technologies at Boddess? A: Our teams have spent the past few months building out our machine learning database of skin types in partnership with leading dermatologists in India. We took thousands of high resolution images of men and women, tagged their skin condition with our proprietary collection of skin metrics (things like hydration, wrinkles, dark spots, skin age etc.).

Social media platforms now serve billions of users by providing convenient means of communication, content sharing and even payment between different users. Due to such convenient and anarchic nature, they have also been used rampantly to promote and conduct business activities between unregistered market participants without paying taxes. Tax authorities worldwide face difficulties in regulating these hidden economy activities by traditional regulatory means. This paper presents a machine learning based Regtech tool for international tax authorities to detect transaction-based tax evasion activities on social media platforms. To build such a tool, we collected a dataset of 58,660 Instagram posts and manually labelled 2,081 sampled posts with multiple properties related to transaction-based tax evasion activities. Based on the dataset, we developed a multi-modal deep neural network to automatically detect suspicious posts. The proposed model combines comments, hashtags and image modalities to produce the final output. As shown by our experiments, the combined model achieved an AUC of 0.808 and F1 score of 0.762, outperforming any single modality models. This tool could help tax authorities to identify audit targets in an efficient and effective manner, and combat social e-commerce tax evasion in scale.