اشترك بالحزمة الذهبية واحصل على وصول غير محدود شمرا أكاديميا
تسجيل مستخدم جديدThe Art of the Meta Stream Protocol: Torrents of Streams
74
0
0.0
(
0
)
نشر من قبل
Christophe De Troyer
تاريخ النشر
2021
مجال البحث
الهندسة المعلوماتية
والبحث باللغة
English
اسأل ChatGPT حول البحث
ﻻ يوجد ملخص باللغة العربية
The rise of streaming libraries such as Akka Stream, Reactive Extensions, and LINQ popularized the declarative functional style of data processing. The stream paradigm offers concise syntax to write down processing pipelines to consume the vast amounts of real-time data available today. These libraries offer the programmer a domain specific language (DSL) embedded in the host language to describe data streams. These libraries however, all suffer from extensibility issues. The semantics of a stream is hard-coded into the DSL language and cannot be changed by the user of the library. We introduce an approach to modify the semantics of a streaming library by means of meta-programming at both run-time and compile-time, and showcase its generality. We show that the expressiveness of the meta-facilities is strong enough to enable push and pull semantics, error handling, parallelism, and operator fusion. We evaluate our work by implementing the identified shortcomings in terms of a novel stream meta-architecture and show that its design and architecture adhere to the design principles of a meta-level architecture. The state of the art offers plenty of choice to programmers regarding reactive stream processing libraries. Expressing reactive systems is otherwise difficult to do in general purpose languages. Extensibility and fine-tuning should be possible in these libraries to ensure a broad variety of applications can be expressed within this single DSL.
قيم البحث
اقرأ أيضاً
We present results from neutral hydrogen (HI) observations of the tip of the Magellanic Stream (MS), obtained with the Arecibo telescope as a part of the on-going survey by the Consortium for Galactic studies with the Arecibo L-band Feed Array. We fi
nd four large-scale, coherent HI streams, extending continously over a length of 20 degrees, each stream possessing different morphology and velocity gradients. The newly discovered streams provide strong support for the tidal model of the MS formation by Connors et al. (2006), which suggested a spatial and kinematic bifurcation of the MS. The observed morphology and kinematics suggest that three of these streams could be interpreted as a 3-way splitting of the main MS filament, while the fourth stream appears much younger and may have originated from the Magellanic Bridge. We find an extensive population of HI clouds at the tip of the MS. Two thirds of clouds have an angular size in the range 3.5--10. We interpret this as being due to thermal instability, which would affect a warm tail of gas trailing through the Galactic halo over a characteristic timescale of a few Myrs to a few hundred Myrs. We show that thermal fragments can survive in the hot halo for a long time, especially if surrounded by a <10^6 K halo gas. If the observed clumpy structure is mainly due to thermal instability, then the tip of the MS is at a distance of ~70 kpc. A significant fraction of HI clouds at the tip of the MS show multi-phase velocity profiles, indicating the co-existence of cooler and warmer gas.
We present high-resolution Magellan/MIKE spectroscopy of 42 red giant stars in seven stellar streams confirmed by the Southern Stellar Stream Spectroscopic Survey (S5): ATLAS, Aliqa Uma, Chenab, Elqui, Indus, Jhelum, and Phoenix. Abundances of 30 ele
ments have been derived from over 10,000 individual line measurements or upper limits using photometric stellar parameters and a standard LTE analysis. This is currently the most extensive set of element abundances for stars in stellar streams. Three streams (ATLAS, Aliqa Uma, and Phoenix) are disrupted metal-poor globular clusters, although only weak evidence is seen for the light element anticorrelations commonly observed in globular clusters. Four streams (Chenab, Elqui, Indus, and Jhelum) are disrupted dwarf galaxies, and their stars display abundance signatures that suggest progenitors with stellar masses ranging from $10^6-10^7 M_odot$. Extensive description is provided for the analysis methods, including the derivation of a new method for including the effect of stellar parameter correlations on each stars abundance and uncertainty. This paper includes data gathered with the 6.5 meter Magellan Telescopes located at Las Campanas Observatory, Chile.
We introduce a new application for inductive logic programming: learning the semantics of programming languages from example evaluations. In this short paper, we explored a simplified task in this domain using the Metagol meta-interpretive learning s
ystem. We highlighted the challenging aspects of this scenario, including abstracting over function symbols, nonterminating examples, and learning non-observed predicates, and proposed extensions to Metagol helpful for overcoming these challenges, which may prove useful in other domains.
We reflect on programming with complicated effects, recalling an undeservingly forgotten alternative to monadic programming and checking to see how well it can actually work in modern functional languages. We adopt and argue the position of factoring
an effectful program into a first-order effectful DSL with a rich, higher-order macro system. Not all programs can be thus factored. Although the approach is not general-purpose, it does admit interesting programs. The effectful DSL is likewise rather problem-specific and lacks general-purpose monadic composition, or even functions. On the upside, it expresses the problem elegantly, is simple to implement and reason about, and lends itself to non-standard interpretations such as code generation (compilation) and abstract interpretation. A specialized DSL is liable to be frequently extended; the experience with the tagless-final style of DSL embedding shown that the DSL evolution can be made painless, with the maximum code reuse. We illustrate the argument on a simple but representative example of a rather complicated effect -- non-determinism, including committed choice. Unexpectedly, it turns out we can write interesting non-deterministic programs in an ML-like language just as naturally and elegantly as in the functional-logic language Curry -- and not only run them but also statically analyze, optimize and compile. The richness of the Meta Language does, in reality, compensate for the simplicity of the effectful DSL. The key idea goes back to the origins of ML as the Meta Language for the Edinburgh LCF theorem prover. Instead of using ML to build theorems, we now build (DSL) programs.
When concept drift is detected during classification in a data stream, a common remedy is to retrain a frameworks classifier. However, this loses useful information if the classifier has learnt the current concept well, and this concept will recur ag
ain in the future. Some frameworks retain and reuse classifiers, but it can be time-consuming to select an appropriate classifier to reuse. These frameworks rarely match the accuracy of state-of-the-art ensemble approaches. For many data stream tasks, speed is important: fast, accurate frameworks are needed for time-dependent applications. We propose the Enhanced Concept Profiling Framework (ECPF), which aims to recognise recurring concepts and reuse a classifier trained previously, enabling accurate classification immediately following a drift. The novelty of ECPF is in how it uses similarity of classifications on new data, between a new classifier and existing classifiers, to quickly identify the best classifier to reuse. It always trains both a new classifier and a reused classifier, and retains the more accurate classifier when concept drift occurs. Finally, it creates a copy of reused classifiers, so a classifier well-suited for a recurring concept will not be impacted by being trained on a different concept. In our experiments, ECPF classifies significantly more accurately than a state-of-the-art classifier reuse framework (Diversity Pool) and a state-of-the-art ensemble technique (Adaptive Random Forest) on synthetic datasets with recurring concepts. It classifies real-world datasets five times faster than Diversity Pool, and six times faster than Adaptive Random Forest and is not significantly less accurate than either.
سجل دخول لتتمكن من نشر تعليقات
التعليقات
جاري جلب التعليقات
سجل دخول لتتمكن من متابعة معايير البحث التي قمت باختيارها
