Browsing by Subject "online learning"
Now showing 1 - 7 of 7
Results Per Page
Sort Options
- ItemOpen AccessAn online learning algorithm for technical trading(2019) Murphy, Nicholas John; Gebbie, TimWe use an adversarial expert based online learning algorithm to learn the optimal parameters required to maximise wealth trading zero-cost portfolio strategies. The learning algorithm is used to determine the relative population dynamics of technical trading strategies that can survive historical back-testing as well as form an overall aggregated portfolio trading strategy from the set of underlying trading strategies implemented on daily and intraday Johannesburg Stock Exchange data. The resulting population time-series are investigated using unsupervised learning for dimensionality reduction and visualisation. A key contribution is that the overall aggregated trading strategies are tested for statistical arbitrage using a novel hypothesis test proposed by Jarrow et al. [31] on both daily sampled and intraday time-scales. The (low frequency) daily sampled strategies fail the arbitrage tests after costs, while the (high frequency) intraday sampled strategies are not falsified as statistical arbitrages after costs. The estimates of trading strategy success, cost of trading and slippage are considered along with an offline benchmark portfolio algorithm for performance comparison. In addition, the algorithms generalisation error is analysed by recovering a probability of back-test overfitting estimate using a nonparametric procedure introduced by Bailey et al. [19]. The work aims to explore and better understand the interplay between different technical trading strategies from a data-informed perspective.
- ItemOpen AccessOnline Non-linear Prediction of Financial Time Series Patterns(2020) da Costa, Joel; Gebbie, TimothyWe consider a mechanistic non-linear machine learning approach to learning signals in financial time series data. A modularised and decoupled algorithm framework is established and is proven on daily sampled closing time-series data for JSE equity markets. The input patterns are based on input data vectors of data windows preprocessed into a sequence of daily, weekly and monthly or quarterly sampled feature measurement changes (log feature fluctuations). The data processing is split into a batch processed step where features are learnt using a Stacked AutoEncoder (SAE) via unsupervised learning, and then both batch and online supervised learning are carried out on Feedforward Neural Networks (FNNs) using these features. The FNN output is a point prediction of measured time-series feature fluctuations (log differenced data) in the future (ex-post). Weight initializations for these networks are implemented with restricted Boltzmann machine pretraining, and variance based initializations. The validity of the FNN backtest results are shown under a rigorous assessment of backtest overfitting using both Combinatorially Symmetrical Cross Validation and Probabilistic and Deflated Sharpe Ratios. Results are further used to develop a view on the phenomenology of financial markets and the value of complex historical data under unstable dynamics.
- ItemOpen AccessPosition Paper: MOOCs(2015) Czerniewicz, Laura; Deacon, Andrew; Fife, Mary-Ann; Small, Janet; Walji, SukainaMassive open online courses (MOOCs) are a flexible and open form of self-directed, online learning designed for mass participation. There are no fees or entry requirements and no formal academic credit is available. While completion rates are low (on average ten per cent) due to varying motivations for enrolling in a MOOC, absolute numbers of participants who complete are usually high. While access to the course material is free, MOOC platform providers often offer certificates of completion at a cost. MOOC platforms provide institutions with cloud-based hosting environments for delivering courses, offering scale and functionality while the institution provides the course material and reputational value. This paper discusses the key aspects of Massive Open Online Courses in a South African educational context.
- ItemOpen AccessResponding to the Initial Challenge of the COVID-19 Pandemic: Analysis of International Responses and Impact in School and Higher Education(Multidisciplinary Digital Publishing Institute, 2022-02-07) Stracke, Christian M.; Burgos, Daniel; Santos-Hermosa, Gema; Bozkurt, Aras; Sharma, Ramesh Chander; Swiatek Cassafieres, Cécile; dos Santos, Andreia Inamorato; Mason, Jon; Ossiannilsson, Ebba; Shon, Jin Gon; Wan, Marian; Obiageli Agbu, Jane-Frances; Farrow, Robert; Karakaya, Özlem; Nerantzi, Chrissi; Ramírez-Montoya, María Soledad; Conole, Grainne; Cox, Glenda; Truong, ViThis paper presents and analyses solutions where open education and open science were utilised to reduce the impact of the COVID-19 pandemic on education. The COVID-19 outbreak and associated lockdowns created huge challenges in school and higher education, demanding sudden responses which aimed to sustain pedagogical quality. Responses have varied from conservative to radically innovative. Universally, the COVID-19 pandemic disrupted and shocked societies worldwide, and education systems were on the front line. The lockdowns largely stopped face-to-face and formal education in almost all countries, and in most cases, distance learning soon became the ‘new normal’. A central challenge concerned sustaining educational visions and ideals in such circumstances. To better understand the state of the art in the educational landscape, we collected case studies from 13 countries during the first year of the pandemic starting on 11 March 2020 (when the World Health Organization declared a pandemic). This paper presents summaries of the full country reports that were collected and describe lessons learned. Our overall aim was to identify good practices and recommendations from the collected case studies that can be taken forward in the future. We categorised the responses on the three generic educational levels (macro, meso and micro) and identified seven key aspects and trends that are valid for all or most countries: (1) formal education at a distance for first time; (2) similar approaches for formal education; (3) missing infrastructure and sharing open educational resources; (4) diverse teaching and learning methods and practices; (5) open education and access to open educational resources; (6) urgent need for professional development and training for teachers and (7) assessing and monitoring learning environments, teachers and students. Finally, we identified key recommendations on how open education and open science can benefit formal education in schools and universities in the future, namely, improved awareness of open educational practices, provision of ICT infrastructure, embracing and sustaining the practice of open access publications and OERs, capacity building for stakeholders and finally encouraging research and development in the area of open education and open science. We found significant evidence for the proposition that open education and open science can support both traditional face-to-face and distance learning.
- ItemOpen AccessStarting with screencasts(2010) Paskevicius, MichaelA short presentation which introduces screencasting: recording the actions and movements performed on a computer screen. Screencasts can be used to demonstrate how to use a particular piece of software/website or to demonstrate a concept using the computer screen much like a classroom blackboard. The recorded videos can then be shared on the web so that anyone can access your lesson. This resource can be used by instructors or anyone interested in creating screencasts.
- ItemOpen AccessStudent Experiences of Emergency Remote Learning and Teaching During COVID-19(2021) Nzala, Athenkosi; Czerniewicz, LauraThis study aimed to explore and understand the University of Cape Town student perceptions and lived experiences of Emergency Remote Teaching and Learning (ERTL) during COVID-19. COVID-19 is a communicable disease instigated by a novel virus (SARS CoV-2 virus). After the inevitable subsequent national lockdown of South Africa, the university placed ERTL measures in place for the second quarter of the first semester to curb the impact of the virus on its students while also enabling learning and teaching activities to continue remotely. ERTL meant that learning and teaching activities were ‘rapidly' shifted from face-to-face learning to remote learning. This study reports on the 707 students who responded to an online survey while engaged in their online courses. The Substitution, Augmentation, Modification, and Redefinition (SAMR) and Andersons' Online Learning Model were used to engage with students on the use of technology that enabled their interaction with lecturers, each other, learning and teaching activities, and other remote learning resources. Understanding the student experiences was achieved through a mixed-method study approach that involved undergraduate and postgraduate students. The Google form online surveys, with both open and closed ended questions with some using the 5-point Likert scale ratings, were distributed using social media platforms and university email system to students in order to collect the data. MAXQDA and Excel software were later utilised to analyse and code the data. Findings for this study indicate that the ERTL experience of the participants during the COVID-19 pandemic presented both opportunities and barriers. Some of the perceived opportunities by students were flexibility and convenience, pedagogical improvements, time saving, self-directed learning (working anytime they want and creating and managing their working schedule), and spending time with family. Interestingly enough, some of these benefits turned out to be challenges for some of the students. Hence, some of the barriers students perceived were distractions, internet connectivity and technical issues, inequitable living and environment conditions, lack of hands-on experience and how this made their degree feel incomplete and difficult, mental health issues, and many other barriers. The disciplinary faculties that experienced most of the obstacles and difficulties associated with ERTL were those whose academic experience depended on practical work in labs and studios or needed software that can only be accessed through labs and would need a specific operating system. The carrying out of this research will help ensure the effectiveness, investment, and continual integration of technology in future programs that involve learning and teaching.
- ItemOpen AccessUsing online social networking for teaching and learning: Facebook use at the University of Cape(2009) Bosch, Tanja EWeb-based learning has made learning content much more freely and instantaneously available to students who can download course notes and readings with a single mouse click. Facebook is one of many Web 2.0 tools – wikis, delicious, YouTube, podcasts – that are listed as having potential applications for teaching and learning. Moreover, it has been argued that the current generation of youth, often described as Net Geners or Digital Natives, may be resistant to traditional methods of teaching and learning. This article explores student use of Facebook at the University of Cape Town, as well as lecturer engagement with students via the new social media. Drawing on a virtual ethnography and qualitative interviews, this article shows that while there are potential positive benefits to using Facebook in teaching and learning, particularly for the development of educational micro-communities, certain challenges, including ICT literacy and uneven access, remain pertinent.