IBM Cloud Research Articles

Development and Application of IoT-Based Emotional Lighting System: Focusing on Emotion Analysis Using Instant Message Data

ABSTRACT Research on human-centric lighting (HCL) is actively being conducted to provide an optimal lighting environment that takes into account the physical and mental states of humans, as well as energy savings and improvements in convenience. This study is a proof-of-concept study that proposes and implements the concept of an IoT-based emotional lighting system using instant message (IM) data. The system we propose performs emotion analysis based on IM data to automatically control the lighting environment of the home according to the emotional state of the occupants when they return home. The lighting system was programmed using Node-RED to extract only the necessary data for emotion analysis from IM data, translate it into English, and perform emotion analysis in IBM Cloud’s Natural Language Understanding (NLU) to adjust the color of the lighting according to the emotional state of the occupants. Based on previous studies that demonstrated the effect of color on emotion regulation, this study selected lighting colors that are appropriate for the emotional state of occupants. The operability of the system was then tested through its implementation in a mock-up studio. Our proposed emotional lighting system, which utilizes IM data, enhances convenience by minimizing the need for occupants to manually intervene. Furthermore, this system can be effortlessly implemented in any personal space and offers scalability due to its seamless integration with other smart devices.

Feedback-based quantum algorithm inspired by counterdiabatic driving

In recent quantum algorithmic developments, a feedback-based approach has shown promise for preparing quantum many-body system ground states and solving combinatorial optimization problems. This method utilizes quantum Lyapunov control to iteratively construct quantum circuits. Here, we propose a substantial enhancement by implementing a protocol that uses ideas from quantum Lyapunov control and the counterdiabatic driving protocol, a key concept from quantum adiabaticity. Our approach introduces an additional control field inspired by counterdiabatic driving. We apply our algorithm to prepare ground states in one-dimensional quantum Ising spin chains. Comprehensive simulations demonstrate a remarkable acceleration in population transfer to low-energy states within a significantly reduced time frame compared to conventional feedback-based quantum algorithms. This acceleration translates to a reduced quantum circuit depth, a critical metric for potential quantum computer implementation. We validate our algorithm on the IBM cloud computer, highlighting its efficacy in expediting quantum computations for many-body systems and combinatorial optimization problems. Published by the American Physical Society 2024

The Role of Cloud Providers in IoT Services

Cloud providers play a key role in IoT services, as they provide the necessary infrastructure, platforms, and services to connect and manage IoT devices and applications. Simply Cloud providers offer These Cloud providers a scalable and flexible computing environment that allows IoT devices to communicate with each other, process data and securely store information. There are several Cloud providers in the market that offer IoT-specific solutions and services such as Amazon Web Services, Microsoft Azure, Google Cloud Platform, IBM Cloud, and Alibaba Cloud. These Cloud Providers offer a range of services specific to IoT, data analytics, and security. This paper compares four Cloud providers, and their advantages and disadvantages. The development of the four Cloud providers and their trends in historical development are also presented with the help of graphs. Finally, a model was created based on three criteria: services, security and big data processing.

Comparative analysis of security models in cloud platforms

This article is dedicated to the importance of data security on cloud platforms, highlighting the subject relevance in an era where threats to information security are becoming increasingly complex and sophisticated. The primary focus is a comparative analysis of security models employed by leading cloud platforms such as Amazon Web Services, Microsoft Azure, Google Cloud Platform, IBM Cloud, and Oracle Cloud. The research covers various crucial security aspects, including data encryption, identity and access management, monitoring mechanisms and incident response. The article sheds light on current trends and technologies in cloud data protection, including the use of artificial intelligence and machine learning to enhance threat detection efficiency and the implementation of confidential computing and blockchain technologies to improve data protection. Issues of compliance with legal requirements and data security standards are discussed, along with recommendations for organisations to optimise information protection in the cloud environment.

A novel quantum Dempster's rule of combination for pattern classification

Dempster's rule of combination (DRC) is widely used for uncertainty reasoning in intelligent information system, which is generalized to complex domain recently. However, as the increase of identification framework elements, the computational complexity of Dempster's rule of combination increases exponentially. To address this issue, we propose a novel quantum Dempster's rule of combination (QDRC) by means of Toffoli gate. The QDRC combination process is completely implemented using quantum circuits. To validate the efficiency of the proposed QDRC, we conduct simulation experiments on IBM and IonQ quantum cloud platforms. Besides, we implement the QDRC algorithm to the real quantum computer of IonQ Harmony. The experiment results show that the result of QDRC is getting more and more accurate as the number of measurements increases. In addition, we propose a quantum multisource information fusion method on the basis of QDRC. Finally, we demonstrate its effectiveness to solve pattern classification problem under several diverse real UCI datasets.

A Comprehensive Comparison of IBM Cloud and Microsoft Azure

The ever-growing demand for scalability, agility, and cost-effectiveness has fuelled cloud computing to the forefront of modern IT infrastructure. According to Precedence Research, the global cloud computing market size was valued at USD 480 billion in 2022 and is expected to hit USD 2297.37 billion by 2032 with a registered CAGR of 17% from 2023 to 2032 [1]. The growth of cloud computing is driven by various factors such as the rapid evolution of technology and the globalization of businesses. Cloud computing is already embraced in various industries like Financial Services, Manufacturing, Healthcare, Retail, Education, and Architecture Engineering & Construction (AEC), where it streamlines operations, enhances collaboration, and revolutionizes processes. Cloud computing holds immense potential, especially in sectors like healthcare, real estate, tourism, and small and medium enterprises, where it will facilitate faster and more informed decision-making processes through data collection and analysis. Additionally, cloud technology is revolutionizing government sectors by driving cost savings, enhancing efficiency, and promoting transparency and citizen engagement. With the increasing number of cloud computing service providers such as IBM, Microsoft, Google, Amazon, and others, it becomes crucial to delve into their technical nuances. Understanding how each provider's offerings align with specific business requirements, technical specifications, and future scalability is paramount for organizations venturing into their cloud journey. The main purpose of this paper is to explore and evaluate the technical dimensions of IBM Cloud and Microsoft Azure and understand its future potential.

Artificial Intelligence Applications Used in On-line Retail in China and Their Relationship to Customer Satisfaction and Loyalty

Artificial Intelligence is the creation of intelligent computers and smart computer algorithms that help machines understand human intelligence (IBM Cloud Education, 2021). Artificial Intelligence is created by analyzing behavior and patterns of big data. Artificial intelligence has existed since the 1950s (Song et al, 2019) but during recent times application of AI in various sectors like healthcare, business, entertainment, education, weather, and geology has picked up momentum. To cite a few well-known examples of Artificial Intelligence technology that are used by almost everyone in our day-to-day activities are advanced Google searches, YouTube recommendations, and Alexa (Song et al, 2019). The advent of technology has brought about innovation and transformation to every aspect of the business world, including the retail industry. The shift in the retail sector becomes obvious when big brick-and-mortar retailers scale down their physical shops and gradually move part of their business online. Consumer attitudes toward online shopping can vary based on demographics, perceived risk, perceived ease of use, and perceived usefulness. Demographic characteristics can further be classified as sex, age, educational qualification, household income, and relationship status, to name a few. In this study, the aim is to identify artificial intelligence applications and their use in online retail by Chinese consumers. A model has been developed to test the relationships between artificial intelligence applications used for online shopping and customer satisfaction and loyalty.

Closing the Cloud Skills Gap with Learner-Centric Performance Support Experiences

Spurred by the urgent need to respond to a growing worldwide technology skills gap, the IBM Cloud Skills Initiative, launched in 2019, today plays an important role in supporting IBM’s commitment to provide people with the skills needed to fill the jobs of tomorrow. The IBM Center for Cloud Training (ICCT) serves as a critical learning resource for helping achieve the initiative’s goal of closing the global cloud skills gap. Through an innovative training program launched in October 2020, ICCT offers Performance Support Experiences (PSEs) designed to motivate and support learners and facilitate their achievement of program credentials and their pursuit of IBM Cloud certification. With PSEs that include online, real-time, and customized study support, ICCT surpassed its stated goal of credentialing participants by the end of 2022. The primary purpose of this article is to describe the components of the program and outcomes achieved and in so doing demonstrate the value of the center’s PSE-based approach to learning.

Abstract 4965: Detection of status of cancer in radiology notes using artificial intelligence

Abstract Introductory Statement: The goal is to use an AI model to replicate the human curator’s response on cancer status in radiology notes of ~20,000 cancer patients. Introduction: MSKCC currently has ~100,000 patients with genomic testing (IMPACT) and continues to accrue more. Clinicians use this genomic data for research but lack clinical structured data to analyze alongside the genomic data. We use a vendor called VASTA Global to manually curate unstructured paragraph text. Depending on the data model, we found that a patient’s full cancer history can take up to 8 hours to curate. Therefore, we want to implement AI to automate the curation of this data to save time and cost. We hope to achieve a faster curation process that allows us to accomplish more than 1 patient a day to catch up to the 100,000 MSK-IMPACT cohort. To decrease the average curation time per patient, we investigated the use of an NLP model to replicate manual curation of the PRISSMM™ ontology field for change in cancer status from radiology reports. Manual curation of radiology reports is time intensive due to curators understanding their cancer status and the volume of scans patients undergo. Methods:A pre-trained Bidirectional Encoder Representations from Transformers (BERT) model is fine-tuned using training data using GPUs in IBM Cloud Pak for Data (CPD) platform. The hyperparameters were adjusted using accuracy and F1 metric of evaluation data. Using one vs. rest approach the model is evaluated on held out test data with results shown in table 1.Table 1 includes class 1-5: Progressing/worsening/enlarging, stable/no change, improving/responding, not stated/indeterminate or mixed. Summary:The weighted AUCROC ~ 0.97, F1 ~ 0.85 and accuracy ~ 93%. These metric scores improve by choosing notes with higher class probabilities only. Conclusion:Due to the above 0.9 calculated accuracy this NLP model is successful in replicating the curated results of cancer status and the next steps will be to run this model on a new cohort of patients not yet curated. Cancer Status Metrics Class Prevalence Precision Recall F1 Score AUCROC Accuracy 1 38.9% 88.7% 87.0% 87.9% 96.5% 90.7% 2 24.7% 89.9% 88.2% 89.1% 98.3% 94.6% 3 15.9% 87.6% 90.7% 89.1% 98.7% 96.5% 4 15.7% 77.0% 72.6% 74.7% 94.4% 92.3% 5 4.8% 68.1% 74.4% 71.1% 97.9% 97.1% Weighted Average 85% 86% 85% 97% 93% Citation Format: Ankur Arya, Andrew Niederhausern, Nadia S. Bahadur, Neil J. Shah, Chelsea Nichols, Avijit Chatterjee, John Philip. Detection of status of cancer in radiology notes using artificial intelligence [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 4965.

The Optimization of a Natural Language Processing Approach for the Automatic Detection of Alzheimer's Disease Using GPT Embeddings.

The development of noninvasive and cost-effective methods of detecting Alzheimer's disease (AD) is essential for its early prevention and mitigation. We optimize the detection of AD using natural language processing (NLP) of spontaneous speech through the use of audio enhancement techniques and novel transcription methodologies. Specifically, we utilized Boll Spectral Subtraction to improve audio fidelity and created transcriptions using state-of-the-art AI services-locally-based Wav2Vec and Whisper, alongside cloud-based IBM Cloud and Rev AI-evaluating their performance against traditional manual transcription methods. Support Vector Machine (SVM) classifiers were then trained and tested using GPT-based embeddings of transcriptions. Our findings revealed that AI-based transcriptions largely outperformed traditional manual ones, with Wav2Vec (enhanced audio) achieving the best accuracy and F-1 score (0.99 for both metrics) for locally-based systems and Rev AI (standard audio) performing the best for cloud-based systems (0.96 for both metrics). Furthermore, this study revealed the detrimental effects of interviewer speech on model performance in addition to the minimal effect of audio enhancement. Based on our findings, current AI transcription and NLP technologies are highly effective at accurately detecting AD with available data but struggle to classify probable AD and mild cognitive impairment (MCI), a prodromal stage of AD, due to a lack of training data, laying the groundwork for the future implementation of an automatic AD detection system.

Automatic Crop Protection from Rain Fall Using IOT and IBM Cloud

With the ongoing fourth revolution, technology has been growing rapidly day by day. One of the most significant and efficient uses of this has been evident in the farming sector. From a basic water sprinkler system to the requirement of Fertilizer for specific crops, the likes of Artificial intelligence and the Internet of things have brought in great changes. One of the issues which need to be addressed is the protection of crops from heavy rainfall which causes significant damage to the crop production and soil yield, also causing huge mental distress to a farmer. We in our proposed model have given a solution to this problem by giving automated as well as manual control. The motivation behind our paper is to keep the crops protected from these heavy precipitations and preserve the same rainwater for future purposes when water is scarce. We use a sensor with a NodeMCU module to ensure the covering of the estimated field and LCD display to show status of Field. This project addresses the critical issue of protecting crops from heavy rainfall through the integration of Internet of Things (IoT) technology and the IBM Cloud platform. By deploying a network of sensors capable of measuring rainfall intensity, soil moisture, temperature, and humidity, real-time data is collected and transmitted to the IBM Cloud for analysis. These measures may include adjusting protective structures, or activating drainage systems. The project aims to improve crop yield and reduce losses caused by adverse weather conditions, making agriculture more sustainable and efficient. The combination of IoT and IBM Cloud empowers farmers with real-time monitoring and control, enhancing their ability to protect valuable crops. This project not only demonstrates the practical application of advanced technologies in agriculture but also provides a foundation for future research and development in this field.

MLLess: Achieving cost efficiency in serverless machine learning training

Function-as-a-Service (FaaS) has raised a growing interest in how to “tame” serverless computing to enable domain-specific use cases such as data-intensive applications and machine learning (ML), to name a few. Recently, several systems have been implemented for training ML models. Certainly, these research articles are significant steps in the correct direction. However, they do not completely answer the nagging question of when serverless ML training can be more cost-effective compared to traditional “serverful” computing. To help in this endeavor, we propose MLLess, a FaaS-based ML training prototype built atop IBM Cloud Functions. To boost cost-efficiency, MLLess implements two innovative optimizations tailored to the traits of serverless computing: on one hand, a significance filter, to make indirect communication more effective, and on the other hand, a scale-in auto-tuner, to reduce cost by benefiting from the FaaS sub-second billing model (often per 100 ms). Our results certify that MLLess can be 15X faster than serverful ML systems [27] at a lower cost for sparse ML models that exhibit fast convergence such as sparse logistic regression and matrix factorization. Furthermore, our results show that MLLess can easily scale out to increasingly large fleets of serverless workers.

Developing an AI-Assisted Educational Chatbot for Radiotherapy Using the IBM Watson Assistant Platform.

Objectives: This study aims to make radiotherapy knowledge regarding healthcare accessible to the general public by developing an AI-powered chatbot. The interactive nature of the chatbot is expected to facilitate better understanding of information on radiotherapy through communication with users. Methods: Using the IBM Watson Assistant platform on IBM Cloud, the chatbot was constructed following a pre-designed flowchart that outlines the conversation flow. This approach ensured the development of the chatbot with a clear mindset and allowed for effective tracking of the conversation. The chatbot is equipped to furnish users with information and quizzes on radiotherapy to assess their understanding of the subject. Results: By adopting a question-and-answer approach, the chatbot can engage in human-like communication with users seeking information about radiotherapy. As some users may feel anxious and struggle to articulate their queries, the chatbot is designed to be user-friendly and reassuring, providing a list of questions for the user to choose from. Feedback on the chatbot's content was mostly positive, despite a few limitations. The chatbot performed well and successfully conveyed knowledge as intended. Conclusions: There is a need to enhance the chatbot's conversation approach to improve user interaction. Including translation capabilities to cater to individuals with different first languages would also be advantageous. Lastly, the newly launched ChatGPT could potentially be developed into a medical chatbot to facilitate knowledge transfer.

Electrochemical Methods in the Cloud: FreiStat, an IoT-Enabled Embedded Potentiostat.

Embedded potentiostats enable electrochemical measurements in the Internet-of-Things (IoT) or other decentralized applications, such as remote environmental monitoring, electrochemical energy systems, and biomedical point-of-care applications. We report on Freiburg's Potentiostat (FreiStat) based on the AD5941 potentiostat circuit from Analog Devices, together with custom firmware, as the key to precise and advanced electrochemical methods. We demonstrated its analytical performance by various cyclic voltammetry measurements, advanced techniques such as differential pulse voltammetry, and a lactate biosensor measurement with currents in the nA range and a resolution of 54 pA. The FreiStat yielded analytical results comparable to benchtop devices and outperformed current commercial embedded potentiostats at significantly lower cost, smaller size, and lower power consumption. Decentralized corrosion analysis by a Tafel plot using the IBM Cloud showed its applicability in a typical IoT scenario. The developed open-source software framework facilitates the integration of electrochemical instrumentation into applications utilizing machine learning and other artificial intelligence. Together with the affordable and highly capable embedded potentiostat, our approach can leverage analytical chemistry toward increasingly important, more widespread and decentralized applications.

Intelligent surveillance systems on the Internet of Things based on secure applications with the IBM cloud platform

The Internet of Things (IoT) is one of the most fast-evolving and most pervasive technological advancements of the last decade. Driven by IoT’s promise for an interconnected world for all people, through a shared Internet-based infrastructure, more and more businesses expand their activities to embrace, competitively, this new concept. More importantly, the ubiquitous, highly granular, and of increasing interconnection and computational capabilities nature of the IoT fabric creates a unique ecosystem for developing powerful, intelligent applications not limited to simply gathering and analyzing environmental and people data. This potential combines the IoT’s physical capabilities with the achievements of the more mature discipline of Artificial Intelligence (AI). However, there is still a large number of businesses with no clear plans of investing in the IoT-plus-AI potential, citing multiple obstacles mainly related to the difficulty of understanding, adapting, and developing advanced AI techniques. Our goal is to show that advanced AI functionalities can, now, be easily implemented and operated through existing IoT platforms, by developing an intelligent, remote, surveillance application based on IBM’s Cloud and Watson platforms as well as the Raspberry Zero W low-cost IoT device.

Quantum computing for heavy quarkonium spectroscopy

We report a first demonstration for the application of quantum computing to heavy quarkonium spectroscopy study. Based on a Cornell-potential model for the heavy quark and antiquark system, we show how this Hamiltonian problem can be formulated and solved with the VQE approach on the IBM cloud quantum computing platform. Errors due to a global depolarizing noise channel are corrected with a zero-noise extrapolation method, resulting in good agreement with the expected value. We also extend the calculation to excited states on a noiseless quantum simulator by orthogonalization with respect to the ground state.

AI and Incidental Findings

AI and Incidental Findings

SECURING DATA AND APPLICATIONS IN CLOUD. CASE STUDY

Security is a principal concern when entrusting an organization’s critical information to geographically dispersed cloud platforms. In addition to the conventional IT system security procedures, designing security into cloud during the software development life cycle can greatly reduce the cloud attack surface. The paper focuses on presenting IBM Cloud security policies and their architecture and conducts a study of the security implementation steps, beginning with user's authentication on the Cloud platform. Hereinafter, the developed experimental application - a monitoring environment system - retrieves temperature, humidity, and pressure values from dedicated sensors, thus using the facilities offered by IBM Cloud platform.

Error mitigation of quantum teleportation on IBM quantum computers

Quantum computers can perform computational tasks beyond the capability of classical computers, such as simulating quantum systems in materials science and chemistry. Quantum teleportation is the transfer of quantum information across distances, relying on entangled states generated by quantum computing. It is becoming a more secure way of sending information, but there is noise in the results. We sought to mitigate the error of quantum teleportation which was simulated on IBM cloud quantum computers. We hypothesized that the noise on all IBM quantum computers could be mitigated with noise-mitigation matrices. We created a quantum teleportation circuit which ran with shots of 500, 1000, 5000, and 8192 for four different qubit states. We studied general error trends in each machine and created two types of noise-mitigation matrices: universal and machine-specific. We then compared the mitigated results of both types of matrices. We found that there was noise for every IBM quantum computer during the trials. The universal noise-mitigation matrix for quantum teleportation for the three tested machines decreased the error for most trials and varied between 1%-5% for most cases after mitigation. The machine-specific noise-mitigation matrix mitigated the error of most machines to just 1-2%, which was a dramatic decrease from unmitigated results (varied between 1%-16% across three machines). The error rates for the machine-specific matrices have less variability than the universal mitigation matrix. We concluded that a universal mitigation matrix could be found for the three machines, but the machine-specific noise mitigation matrices were able to achieve more accurate results.

Quantum algorithm for learning secret strings and its experimental demonstration

In this paper, we consider the secret-string-learning problem in the teacher–student setting: the teacher has a secret string s∈{0,1}n, and the student wants to learn the secret s by question–answer interactions with the teacher, where at each time, the student can ask the teacher with a pair (x,q)∈{0,1}n×{0,1,…,n−1} and the teacher returns a bit given by the oracle fs(x,q) that indicates whether the length of the longest common prefix of s and x is greater than q or not. Our contributions are as follows.(i) We prove that any classical deterministic algorithm needs at least n queries to the oracle fs to learn the n-bit secret string s in both the worst case and the average case, and also present an optimal classical deterministic algorithm learning any s using n queries.(ii) We propose a quantum algorithm learning the n-bit secret string s with certainty using n/2 queries to the oracle fs, thus proving a double speedup over classical counterparts.(iii) Experimental demonstrations of our quantum algorithm on the IBM cloud quantum computer are presented, with the average success probabilities of 85.3% and 82.5% for all cases with n=2 and n=3, respectively.