Sequential Learning of Principal Curves: Summarizing Data Streams on the Fly

When confronted with massive data streams, summarizing data with dimension reduction methods such as PCA raises theoretical and algorithmic pitfalls. A principal curve acts as a nonlinear generalization of PCA, and the present paper proposes a novel algorithm to automatically and sequentially learn principal curves from data streams. We show that our procedure is supported by regret bounds with optimal sublinear remainder terms. A greedy local search implementation (called slpc, for sequential learning principal curves) that incorporates both sleeping experts and multi-armed bandit ingredients is presented, along with its regret computation and performance on synthetic and real-life data.

Age affects pigeons’ (Columba livia) memory capacity but not representation of serial order during a locomotor sequential-learning task

Aging affects individuals of every species, with sometimes detrimental effects on memory and cognition. The simultaneous-chaining task, a sequential-learning task, requires subjects to select items in a predetermined sequence, putting demands on memory and cognitive processing capacity. It is thus a useful tool to investigate age-related differences in these domains. Pigeons of three age groups (young, adult and aged) completed a locomotor adaptation of the task, learning a list of four items. Training began by presenting only the first item; additional items were added, one at a time, once previous items were reliably selected in their correct order. Although memory capacity declined noticeably with age, not all aged pigeons showed impairments compared to younger pigeons, suggesting that inter-individual variability emerged with age. During a subsequent free-recall memory test in the absence of reinforcement, when all trained items were presented alongside novel distractor items, most pigeons did not reproduce the trained sequence. During a further forced-choice test, when pigeons were given a choice between only two of the trained items, all three age groups showed evidence of an understanding of the ordinal relationship between items by choosing the earlier item, indicating that complex cognitive processing, unlike memory capacity, remained unaffected by age.

E-Learning and Motor Relearning Programme in Physical Education

Today's readers engage in compelling, moving, customized, and customized content. This need is met by an online learning style, where students can study voluntarily and on their own. The effects of digital integration are also evident in the education sector and have contributed to significant changes in the way education is taught and consumed. In the context of new normal e-learning is going on with fast paces where the physical component is lacking. The absence of physical work out may affect the performance of the students which needs motor relearning in the next normal scenario. The present article deals with e-learning with motor relearning programmer with emphasis on implicit, explicit and sequential learning.

Sequential learning to accelerate discovery of alkali-activated binders

Alkali-activated binders (AAB) can provide a clean alternative to conventional cement in terms of CO2 emissions. However, as yet there are no sufficiently accurate material models to effectively predict the AAB properties, thus making optimal mix design highly costly and reducing the attractiveness of such binders. This work adopts sequential learning (SL) in high-dimensional material spaces (consisting of composition and processing data) to find AABs that exhibit desired properties. The SL approach combines machine learning models and feedback from real experiments. For this purpose, 131 data points were collected from different publications. The data sources are described in detail, and the differences between the binders are discussed. The sought-after target property is the compressive strength of the binders after 28 days. The success is benchmarked in terms of the number of experiments required to find materials with the desired strength. The influence of some constraints was systematically analyzed, e.g., the possibility to parallelize the experiments, the influence of the chosen algorithm and the size of the training data set. The results show the advantage of SL, i.e., the amount of data required can potentially be reduced by at least one order of magnitude compared to traditional machine learning models, while at the same time exploiting highly complex information. This brings applications in laboratory practice within reach.