Wing dimorphisms have long served as models for examining the ecological and evolutionary tradeoffs associated with alternative phenotypes. Here, we investigated genetic cause of pea aphid (Acyrthosiphon pisum) male wing dimorphism, wherein males exhibit one two morphologies that differ in correlated traits include presence or absence wings. We mapped this trait difference to a single genomic region and, using third generation, long-read sequencing, identified 120 kb insertion wingless...

Locality analysis is important since accessing memory much slower than computing. Compile-time locality can provide detailed program-level feedback for compilers or runtime systems faster trace-based analysis.

Modeling prosodic rhythm is of great importance for both speech synthesis and understanding, it requires a large enough corpus with precise boundary labels. This paper proposes maximum entropy (ME) based hierarchical model, which utilizes text acoustic features, to automatically label Mandarin boundaries. Results comparative experiments show that, the task detection, ME model obviously outperforms classification regression tree (CART), bottom-up framework also significantly superior flat...

Many users with severe motor impairments, such as quadriplegics, interact computers using an indirect selection technique called switch access scanning. Switch scanning allows for iteratively selecting input from a set of options single and which replaces the use keyboard or mouse, they may be unable to use. Navigating avatar in 3D virtual world existing systems is slow erroneous because these interfaces are non-linear requires players provide continuous (holding key) mixed inputs two more...

Recent years have seen rising research in logic synthesis recipe generation to improve the Quality-of-Result (QoR). However, existing approaches typically low efficiency and are stuck at local optima. In this work, we propose a optimization framework, AlphaSyn, that incorporates domain-specific Monte Carlo tree search (MCTS) algorithm. AlphaSyn enables exploration across entire space while optimizing sampling points utilization. We further develop synthesis-specific upper confidence bound...

Homograph disambiguation is the core issue of graphemetophoneme conversion in Mandarin Text-to-Speech system. In this paper, a hybrid algorithm called tree-guided transformation-based learning (TTBL), which combines decision tree with (TBL), proposed to resolve homograph ambiguity. It can automatically generate templates, thereby avoiding manually summarizing time-consuming and laborious conventional TBL. addition, paper evaluates various keyword selection approaches different domains....

The WISTON system is a large corpus based TTS with the unit selection method. text analysis part of this contains pre-processing, word segmentation, POS tagging, phonetic transcription and prosody structure prediction. information (duration, F0, energy) predicted by CART model input context information. In model, we use mutual constraint as concatenation costs for path searching while F0s, durations energies are used to get target costs. spectrum smoothing method also speech generation....

Federated learning is the distributed machine framework that enables collaborative training across multiple parties while ensuring data privacy. Practical adaptation of XGBoost, state-of-the-art tree boosting framework, to federated remains limited due high cost incurred by conventional privacy-preserving methods. To address problem, we propose two variants XGBoost with privacy guarantee: FedXGBoost-SMM and FedXGBoost-LDP. Our first protocol deploys enhanced secure matrix multiplication...

Locality analysis is important since accessing memory much slower than computing. Compile-time locality can provide detailed program-level feedback for compilers or runtime systems faster trace-based analysis. In this paper, we describe a new approach to based on static parallel sampling. A compiler analyzes loop-based code and generates sampler which run measure locality. Our predict precise cache line granularity miss ratio curves complex loops with non-linear array references even...

Cache management is important in exploiting locality and reducing data movement. This article studies a new type of programmable cache called the lease cache. By assigning leases, software exerts primary control on when how long stays Previous work has shown an optimal solution for ideal develops evaluates set practical solutions physical emulated FPGA with full suite PolyBench benchmarks. Compared to automatic caching, programming can further reduce movement by 10% over 60% size 16 times...

This paper investigates the stabilization and optimization problems for a group of identically linear agents with undirected interaction topology. It is shown that distributed control law based on local measurements relative information exchanged from neighboring can be designed each agent to enable states stabilized. Furthermore, due use parametric Lyapunov approach, guarantees not only performance at network level but also convergence rate agents. Finally, simulation example provided...

Two common techniques in efficient caching are associativity and sub-block granularity. This short paper presents a parameterized composable model for each of the two techniques. It shows how new models more general, accurate or than previous modeling solutions either technique, they can be used together to cache implemented with both techniques, i.e. set associative cache.

Cache in multicore machines is often shared, and the cache performance depends on how memory accesses belonging to different programs interleave with one another. The full range of possibilities includes all possible interleavings, which are too numerous be studied by experiments for any mix non-trivial programs.

Data movement is a common performance bottleneck, and its chief remedy caching. Traditional cache management transparent to the workload: data that should be kept in are determined by recency information only, while program information, i.e., future reuses, not communicated cache. This has changed new design named Lease Cache . The control passed lease compiler technique called Compiler Assigned Reference (CARL). collects reuse interval distribution for each reference uses it compute assign...

Lease caching is a new technique that provides greater control of the cache than what allowed in conventional caches. The simplest uniform lease (UL), which means all leases are identical length. UL prescriptive and based on allocation. In comparison, reactive replacement. They represent two fundamentally different approaches to management.

Feature Extraction (FE) based on Principal Component Analysis (PCA) can effectively improve classification results by reducing the interference among features. However, such a good method has not been employed in previous studies of Incremental Attribute Learning (IAL), novel machine learning strategy, where features are gradually trained one order to remove and results. This study proposed preprocessing for neural IAL algorithm feature extraction with PCA. Experimental show that this...

Consensus control in multi-agent systems has received significant attention and practical implementation across various domains. However, managing consensus under unknown dynamics remains a challenge for design due to system uncertainties environmental disturbances. This paper presents novel learning-based distributed law, augmented by an auxiliary dynamics. Gaussian processes are harnessed compensate the components of system. For continuous enhancement predictive performance process model,...