Partition Tree Research Articles

Comparative study of machine learning approaches integrated with genetic algorithm for IVF success prediction.

IVF is a widely-used assisted reproductive technology with a consistent success rate of around 30%, and improving this rate is crucial due to emotional, financial, and health-related implications for infertile couples. This study aimed to develop a model for predicting IVF outcome by comparing five machine-learning techniques. The research approached five prominent machine learning algorithms, including Random Forest, Artificial Neural Network (ANN), Support Vector Machine (SVM), Recursive Partitioning and Regression Trees (RPART), and AdaBoost, in the context of IVF success prediction. The study also incorporated GA as a feature selection method to enhance the predictive models' robustness. Findings demonstrate that AdaBoost, particularly when combined with GA feature selection, achieved the highest accuracy rate of 89.8%. Using GA, Random Forest also demonstrated strong performance, achieving an accuracy rate of 87.4%. Genetic Algorithm significantly improved the performance of all classifiers, emphasizing the importance of feature selection. Ten crucial features, including female age, AMH, endometrial thickness, sperm count, and various indicators of oocyte and embryo quality, were identified as key determinants of IVF success. These findings underscore the potential of machine learning and feature selection techniques to assist IVF clinicians in providing more accurate predictions, enabling tailored treatment plans for each patient. Future research and validation can further enhance the practicality and reliability of these predictive models in clinical IVF practice.

PyPLIF HIPPOS-aided Construction and Retrospective Validation of Structure-Based Virtual Screening Protocol Targeting VEGFR2

Recently, the discovery of small molecules as inhibitors for vascular endothelial growth factor receptor 2 (VEGFR2) is of timely interest, especially in the area of ocular neovascular diseases. On the other hand, PyPLIF HIPPOS in combination with machine learning techniques has been reported to increase the prediction quality of structure-based virtual screening (SBVS) protocols. The original version of PyPLIF has served in the development of an SBVS protocol that successfully identified novel chalcone derivatives and short peptides as potent inhibitors for acetylcholinesterase. In this short communication, construction and retrospective validation of an SBVS protocol employing PyPLIF HIPPOS targeting VEGFR2 are presented to make it publicly available. The retrospective validation employed 409 active compounds and 24,950 decoys from the enhanced version of the directory of useful decoys. All compounds were docked independently 3 times using AutoDock Vina followed by the identification of the protein-ligand interaction fingerprints (PLIF) employing PyPLIF HIPPOS. The derived ensemble PLIF descriptors were then used in the decision tree construction using a machine-learning technique called recursive partitioning and regression trees. The best decision was then incorporated in the SBVS protocol. The F-measure and enrichment factor values of the SBVS protocol were 0.387 and 76.879, respectively. Hence, the SBVS protocol is readily available to screen small molecules or short peptides.

Leveraging occupancy map to accelerate video-based point cloud compression

Video-based Point Cloud Compression enables point cloud streaming over the internet by converting dynamic 3D point clouds to 2D geometry and attribute videos, which are then compressed using 2D video codecs like H.266/VVC. However, the complex encoding process of H.266/VVC, such as the quadtree with nested multi-type tree (QTMT) partition, greatly hinders the practical application of V-PCC. To address this issue, we propose a fast CU partition method dedicated to V-PCC to accelerate the coding process. Specifically, we classify coding units (CUs) of projected images into three categories based on the occupancy map of a point cloud: unoccupied, partially occupied, and fully occupied. Subsequently, we employ either statistic-based rules or machine-learning models to manage the partition of each category. For unoccupied CUs, we terminate the partition directly; for partially occupied CUs with explicit directions, we selectively skip certain partition candidates; for the remaining CUs (partially occupied CUs with complex directions and fully occupied CUs), we train an edge-driven LightGBM model to predict the partition probability of each partition candidate automatically. Only partitions with high probabilities are retained for further Rate–Distortion (R–D) decisions. Comprehensive experiments demonstrate the superior performance of our proposed method: under the V-PCC common test conditions, our method reduces encoding time by 52% and 44% in geometry and attribute, respectively, while incurring only 0.68% (0.66%) BD-Rate loss in D1 (D2) measurements and 0.79% (luma) BD-Rate loss in attribute, significantly surpassing state-of-the-art works.

Image Compression Based on DWT Using EZW/SPIHT Encoders

this paper reviews recent research on image compression techniques utilizing the Embedded Zero-Tree Wavelet (EZW) and Set Partitioning in Hierarchical Trees (SPIHT) algorithms. The study addresses the question of which algorithm, EZW or SPIHT, provides superior compression performance for various image types. The research method involves a comprehensive analysis of published research papers that employed these algorithms for image compression. The findings indicate that SPIHT generally outperforms EZW in terms of compression ratio, computational efficiency, and image quality, making it a more versatile and effective choice for various image compression applications.

Schistura sonarengaensis, a new species of cave-dwelling loach (Teleostei: Nemacheilidae) from Meghalaya, northeast India.

A new species of nemacheilid loach, Schistura sonarengaensis sp. nov., is described from three cave-dwelling populations (Barak-Surma-Meghna drainage) in the South Garo Hills district of Meghalaya, India. The new species possesses prominent eyes but is easily distinguished from all the congeners of the genus Schistura from Barak-Surma-Meghna and adjacent rivers drainages of northeast Indian (except S. syngkai) in having 13-26 vertically elongated to circular mid-lateral black blotches (brownish in life) overlayered on a grayish-black mid-lateral stripe on a dull white or pale-beige (golden brown in life) body. However, it can be easily distinguished from S. syngkai in having a complete (vs. incomplete) lateral line with more 72-89 (vs. 19-42) lateral-line pored scales, greater pre-dorsal length (48.5-53.1 vs. 41.9-44.1 %SL), a wider body at dorsal-fin origin (11.3-16.7 vs. 9.4-10.3 %SL), greater dorsal (18.1-21.1 vs. 15.1-17.0 %SL) and lateral (20.9-24.1 vs. 17.4-18.9 %SL) head length, a wider head (14.5-18.5 vs. 11.6-13.3 %SL), and moderately forked (vs. emarginated) caudal fin. Further, molecular analysis confirms the distinctiveness of S. sonarengaensis sp. nov. from its congeners found in northeast India by significant divergences with uncorrected genetic distance ranging from 3.7% to 17.3% in the mitochondrial cytochrome c oxidase subunit I (COI) gene dataset. The phylogenetic position of the new species with its sister species was evaluated using maximum likelihood and Bayesian analysis. The species delimitation approaches assemble species by automatic partitioning (ASAP) and Poisson tree processes (PTP) utilized for testing species assignments consistently identified our test group as a distinct species from its sister species. Although the new species lacks typical morphological adaptations usually associated with a subterranean life, such as complete absence (or vestigial presence) of eyes and pigmentation, it exhibits a reduction of pigmentation when compared to the epigean congeners.

Early quadtree with nested multitype tree partitioning algorithm based on convolution neural network for the versatile video coding standard

Early quadtree with nested multitype tree partitioning algorithm based on convolution neural network for the versatile video coding standard

Secure Range-Searching Using Copy-And-Recurse

Range searching is the problem of preprocessing a set of points P, such that given a query range gamma we can efficiently compute some function f(P cap gamma). For example, in a 1 dimensional range counting query, P is a set of numbers, gamma is a segment and we need to count how many numbers of P are in gamma. In higher dimensions, P is a set of d dimensional points and the query range is some volume in R^d. In general, we want to compute more than just counting, for example, the average of P cap gamma. Range searching has applications in databases where some SELECT queries can be translated to range queries. It had received a lot of attention in computational geometry where a data structure called partition tree was shown to solve range queries in time sub-linear in |P| using space only linear in |P|. In this paper we consider partition trees under FHE where we answer range queries without learning the value of the points or the parameters of the range. We show how partition trees can be securely traversed with O(t n^{1-1/d+epsilon} + n^{1+epsilon}) operations, where n=|P|, t is the number of operations needed to compare to gamma and epsilon>0 is a parameter. When the ranges are axis-parallel hyper-boxes the running time is O(t n^epsilon + n log^{d-1} n). As far as we know, this is the first non-trivial bound on range searching under FHE and it improves over the naive solution that needs O(t n) operations. Our algorithms are independent of the encryption scheme but as an example we implemented them using the CKKS FHE scheme. Our experiments show that for databases of sizes 2^{23} and 2^{25}, our algorithms run x2.8 and x4.7 (respectively) faster than the naive algorithm. The improvement of our algorithm comes from a method we call copy-and-recurse. With it we efficiently traverse a r-ary tree (where each inner node has r children) that also has the property that at most xi of them need to be recursed into when traversing the tree. We believe this method is interesting in its own and can be used to improve traversals in other tree-like structures.

Predictions of Colloidal Molecular Aggregation Using AI/ML Models.

To facilitate the triage of hits from small molecule screens, we have used various AI/ML techniques and experimentally observed data sets to build models aimed at predicting colloidal aggregation of small organic molecules in aqueous solution. We have found that Naïve Bayesian and deep neural networks outperform logistic regression, recursive partitioning tree, support vector machine, and random forest techniques by having the lowest balanced error rate (BER) for the test set. Derived predictive classification models consistently and successfully discriminated aggregator molecules from nonaggregator hits. An analysis of molecular descriptors in favor of colloidal aggregation confirms previous observations (hydrophobicity, molecular weight, and solubility) in addition to undescribed molecular descriptors such as the fraction of sp3 carbon atoms (Fsp3), and electrotopological state of hydroxyl groups (ES_Sum_sOH). Naïve Bayesian modeling and scaffold tree analysis have revealed chemical features/scaffolds contributing the most to colloidal aggregation and nonaggregation, respectively. These results highlight the importance of scaffolds with high Fsp3 values in promoting nonaggregation. Matched molecular pair analysis (MMPA) has also deciphered context-dependent substitutions, which can be used to design nonaggregator molecules. We found that most matched molecular pairs have a neutral effect on aggregation propensity. We have prospectively applied our predictive models to assist in chemical library triage for optimal plate selection diversity and purchase for high throughput screening (HTS) in drug discovery projects.

Основанный на генетическом подходе алгоритм внутрикодирования для H.266/VVC

This paper presents a genetic approach for optimizing intra coding in H.266/VVC. The proposed algorithm efficiently selects coding tools and Multi-Type Tree (MTT) partitions to achieve a balance between encoding time and video quality. The fitness evaluation function, which combines perceptual metrics and coding efficiency metrics, is used to assess the quality of each candidate solution. The results demonstrate a significant reduction in encoding time without compromising video quality. The proposed algorithm selects coding tools from a set of available tools in H.266/VVC. These tools include intra prediction modes, transform units, quantization parameters, and entropy coding modes. The MTT partitioning scheme includes four types of partitions: quadtree, binary tree, ternary tree, and quad-binary tree. Perceptual metrics are used to evaluate the visual quality of the encoded video. Coding efficiency metrics are used to evaluate the coding efficiency of the encoded video. The fitness evaluation function combines perceptual metrics and coding efficiency metrics to assess the quality of each candidate solution.

DET-LSH: A Locality-Sensitive Hashing Scheme with Dynamic Encoding Tree for Approximate Nearest Neighbor Search

Locality-sensitive hashing (LSH) is a well-known solution for approximate nearest neighbor (ANN) search in high-dimensional spaces due to its robust theoretical guarantee on query accuracy. Traditional LSH-based methods mainly focus on improving the efficiency and accuracy of the query phase by designing different query strategies, but pay little attention to improving the efficiency of the indexing phase. They typically fine-tune existing data-oriented partitioning trees to index data points and support their query strategies. However, their strategy to directly partition the multi-dimensional space is time-consuming, and performance degrades as the space dimensionality increases. In this paper, we design an encoding-based tree called Dynamic Encoding Tree (DE-Tree) to improve the indexing efficiency and support efficient range queries based on Euclidean distance. Based on DE-Tree, we propose a novel LSH scheme called DET-LSH. DET-LSH adopts a novel query strategy, which performs range queries in multiple independent index DE-Trees to reduce the probability of missing exact NN points, thereby improving the query accuracy. Our theoretical studies show that DET-LSH enjoys probabilistic guarantees on query accuracy. Extensive experiments on real-world datasets demonstrate the superiority of DET-LSH over the state-of-the-art LSH-based methods on both efficiency and accuracy. While achieving better query accuracy than competitors, DET-LSH achieves up to 6x speedup in indexing time and 2x speedup in query time over the state-of-the-art LSH-based methods.

A new hybrid GPU-CPU sparse LDLT factorization algorithm with GPU and CPU factorizing concurrently

This paper proposes a new task assignment scheme for sparse LDLT factorization on a hybrid GPU-CPU platform, and correspondingly an efficient supernodal algorithm, GPU Node and CPU Pipelining (GNCP). GNCP assigns a large number of weakly coupled tasks to GPU and CPU respectively so that GPU and CPU can execute different factorization tasks concurrently without explicit synchronization, which has not been achieved by previous researches. More precisely, based on the number of CPU threads, the global memory of GPU as well the size of the matrix, we introduce the concept of the truncation level in the partition tree built by the multi-level graph partitioning. The sum of FLOP associated with decoupled nodes in the truncation level accounts for about 65% of the whole computations in tested numerical examples. GNCP assigns the factorization tasks of each node in the truncation level to GPU. The data of each node in the truncation level are copied to the global memory of GPU and factorized entirely and independently by GPU one-by-one. Once factorization of any node in the truncation level has been completed, the cross-factorizations of this node to its ancestors and the following related operations are decoupled from other GPU tasks. These tasks are assigned to CPU threads with a reentrant pipelining strategy, accounting for about 35% of the whole calculations. Using this novel design, CPU and GPU are working concurrently and a large part of factorization is performed in overlapped time. Numerical tests for matrices from SuiteSparse matrix collection show higher performance than the well-known hybrid GPU-CPU solver CHOLMOD.

Generation of PN-Model Synthesis Programs with Restrictions

The paper proposes a modified algorithm for generating programs for the synthesis of network models expressed in terms of Petri nets [8]. The algorithm is based on the use of boundedly growing strings and the ability to generate them in lexicographic order. To reduce the number of possible synthesis programs, the introduction of limiting conditions imposed on the set partition tree is considered. Taking into account the imposed restrictions, a new tree of smaller dimension is built, which also preserves the lexicographic order of the limitedly growing strings it contains. A comparative assessment of the effectiveness of a new algorithm for generating boundedly growing strings with restrictions and a standard algorithm without restrictions, which is described by D. Knuth, is given. The proposed algorithm can be used to solve problems of synthesizing models of complex computational structures, building parallel systems, parallelizing algorithms and solving other problems related to the creation of efficient distributed systems.

Ten years of hand hygiene excellence: a summary of outcomes, and a comparison of indicators, from award-winning hospitals worldwide.

Hand hygiene is a crucial measure for the prevention of healthcare-associated infections (HAIs). The Hand Hygiene Excellence Award (HHEA) is an international programme acknowledging healthcare facilities for their leadership in implementing hand hygiene improvement programmes, including the World Health Organisation's Multimodal Improvement Strategy. This study aimed at summarising the results of the HHEA campaign between 2010 and 2021 and investigating the relationship between different hand hygiene parameters based on data from participating healthcare facilities. A retrospective analysis was performed on datasets from HHEA forms, including data on hand hygiene compliance, alcohol-based handrub (ABHR) consumption, and Hand Hygiene Self-Assessment Framework (HHSAF) scores. Descriptive statistics were reported for each variable. The correlation between variables was inspected through Kendall's test, while possible non-linear relationships between hand hygiene compliance, ABHR consumption and HHSAF scores were sought through the Locally Estimated Scatterplot Smoothing or logistic regression models. A tree-structured partitioning model was developed to further confirm the obtained findings. Ninety-seven healthcare facilities from 28 countries in three world regions (Asia-Pacific, Europe, Latin America) were awarded the HHEA and thus included in the analysis. HHSAF scores indicated an advanced hand hygiene promotion level (median 445 points, IQR 395-480). System change (100 [95-100] points) and institutional safety climate (85 [70-95] points) showed the highest and lowest score, respectively. In most cases, hand hygiene compliance was above 70%, with heterogeneity between countries. ABHR consumption above 20 millilitres per patient-day (ml/PD) was widely reported, with overall increasing trends. HHSAF scores were positively correlated with hand hygiene compliance (τ = 0.211, p = 0.007). We observed a positive correlation between compliance rates and ABHR consumption (τ = 0.193, p < 0.001), although the average predicted consumption was stable around 55-60ml/PD for compliance rates above 80-85%. Logistic regression and partitioning tree analyses revealed that higher HHSAF scores were more likely in the high-ABHR consumption group at cut-offs around 57-59ml/PD. Ten years after its inception, the HHEA proves to be a valuable hand hygiene improvement programme in healthcare facilities worldwide. Consistent results were provided by the different hand hygiene indicators and the HHSAF score represents a valuable proxy measure of hand hygiene compliance.

Carbon budget at the individual-tree scale: dominant Eucalyptus trees partition less carbon belowground.

Large trees in plantations generally produce more wood per unit of resource use than small trees. Two processes may account for this pattern: greater photosynthetic resource use efficiency or greater partitioning of carbon to wood production. We estimated gross primary production (GPP) at the individual scale by combining transpiration with photosynthetic water-use efficiency of Eucalyptus trees. Aboveground production fluxes were estimated using allometric equations and modeled respiration; total belowground carbon fluxes (TBCF) were estimated by subtracting aboveground fluxes from GPP. Partitioning was estimated by dividing component fluxes by GPP. Dominant trees produced almost three times as much wood as suppressed trees. They used 25 ± 10% (mean ± SD) of their photosynthates for wood production, whereas suppressed trees only used 12 ± 2%. By contrast, dominant trees used 27 ± 19% of their photosynthate belowground, whereas suppressed trees used 58 ± 5%. Intermediate trees lay between these extremes. Photosynthetic water-use efficiency of dominant trees was c. 13% greater than the efficiency of suppressed trees. Suppressed trees used more than twice as much of their photosynthate belowground and less than half as much aboveground compared with dominant trees. Differences in carbon partitioning were much greater than differences in GPP or photosynthetic water-use efficiency.

VLSI design for adjustable compression rate in lossless/lossy compression of EEG signal

In the realm of remote health monitoring, the compression of electroencephalography (EEG) signals at the edge holds significant importance. This paper presents a VLSI structure capable of adjustable compression rates to achieve both lossless and lossy compression of EEG data, catering to environments requiring flexible adjustment of compression types and rates. The architecture integrates two-dimensional discrete wavelet transform (2D-DWT) and the Set Partitioning in Hierarchical Trees (SPIHT). Notably, the Haar wavelet lifting structure is employed in 2D-DWT, and the SPIHT scanning sequence is optimized for enhanced resource efficiency. The proposed algorithm is validated using the CHB-MIT Scalp EEG Database, demonstrating compression ratios of ≥1.95 for normal EEG and ≥1.69 for epileptic EEG. Through synthesis and layout design in the 130 nm CMOS technology, the resulting circuit achieves a maximum clock frequency of 57 MHz and occupies an area of 81 kμm2, with an average power consumption of less than 0.71nj per data for 16-bit data. These findings affirm the efficacy of the proposed VLSI structure for achieving efficient lossless/lossy compression of edge EEG data.

Development of a hard cider flavor wheel using free word sorting and multivariate statistical techniques

AbstractFlavor wheels are visual tools built from standardized sensory lexicons that are used in many different industries to improve communication, marketing, and quality discrimination among products. To date, flavor wheels have been developed for spices, coffee, wine, beer, and many other foods and beverages—yet no flavor wheel has been constructed for hard cider. While there is no single established method for constructed flavor wheels, most approaches are based on sensory descriptive analysis (DA) and free word sorting activities that investigate the semantic similarity and dissimilarity of descriptive terms. This research study utilized multiple DA studies for the generation of a sensory lexicon, followed by independent word sorting tasks with cider industry professionals (N = 40) and untrained consumers (N = 58) to establish two flavor wheels that are broadly understandable to a variety of industry stakeholders. Based on the results of DISTATIS and additive tree partitioning, this research showcases a workflow for developing and refining flavor wheels that incorporate both stakeholder and researcher input and can be built upon by other analysts.Practical ApplicationsThis research demonstrates an accessible methodology for developing flavor wheels that incorporates input from diverse parties and considers the semantic associations of terms used by cider consumers and industry professionals. The proposed methodology is a useful framework for other sensory scientists seeking to efficiently develop flavor wheels for multiple audiences. As a demonstration, this research also delivers two flavor wheels—one that highlights discrepancies in how industry consumers and producers semantically understand sensory experiences compared to trained sensory personnel, and a second wheel that showcases how hard cider can more objectively be described. Together, these flavor wheels are tools for improving sensory communication, education, and marketing in the US hard cider industry.

Sample-and-Bound for Non-convex Optimization

Standard approaches for global optimization of non-convex functions, such as branch-and-bound, maintain partition trees to systematically prune the domain. The tree size grows exponentially in the number of dimensions. We propose new sampling-based methods for non-convex optimization that adapts Monte Carlo Tree Search (MCTS) to improve efficiency. Instead of the standard use of visitation count in Upper Confidence Bounds, we utilize numerical overapproximations of the objective as an uncertainty metric, and also take into account of sampled estimates of first-order and second-order information. The Monte Carlo tree in our approach avoids the usual fixed combinatorial patterns in growing the tree, and aggressively zooms into the promising regions, while still balancing exploration and exploitation. We evaluate the proposed algorithms on high-dimensional non-convex optimization benchmarks against competitive baselines and analyze the effects of the hyper parameters.

Machine learning-based algorithms applied to drug prescriptions and other healthcare services in the Sicilian claims database to identify acromegaly as a model for the earlier diagnosis of rare diseases

Acromegaly is a rare disease characterized by a diagnostic delay ranging from 5 to 10 years from the symptoms’ onset. The aim of this study was to develop and internally validate machine-learning algorithms to identify a combination of variables for the early diagnosis of acromegaly. This retrospective population-based study was conducted between 2011 and 2018 using data from the claims databases of Sicily Region, in Southern Italy. To identify combinations of potential predictors of acromegaly diagnosis, conditional and unconditional penalized multivariable logistic regression models and three machine learning algorithms (i.e., the Recursive Partitioning and Regression Tree, the Random Forest and the Support Vector Machine) were used, and their performance was evaluated. The random forest (RF) algorithm achieved the highest Area under the ROC Curve value of 0.83 (95% CI 0.79–0.87). The sensitivity in the test set, computed at the optimal threshold of predicted probabilities, ranged from 28% for the unconditional logistic regression model to 69% for the RF. Overall, the only diagnosis predictor selected by all five models and algorithms was the number of immunosuppressants-related pharmacy claims. The other predictors selected by at least two models were eventually combined in an unconditional logistic regression to develop a meta-score that achieved an acceptable discrimination accuracy (AUC = 0.71, 95% CI 0.66–0.75). Findings of this study showed that data-driven machine learning algorithms may play a role in supporting the early diagnosis of rare diseases such as acromegaly.

Myocarditis-A Helpful Algorithm to Overcome Diagnostic Challenges in the Pediatric Population.

This study was designed to investigate clinical differences between pediatric patients who presented with chest pain, tachycardia, and/or tachypnea who subsequently were or were not diagnosed with myocarditis. The results were used to develop a decision tree to aid in rapid diagnosis of pediatric myocarditis. A retrospective case-control study was performed using the electronic medical records of children aged 0 to 18 years between the years 2003 and 2020 with a complaint of chest pain, tachycardia, and/or tachypnea. Patients included in the study were those diagnosed with myocarditis and those with suspected myocarditis, which was ultimately ruled out. Demographic and clinical differences between the research groups were analyzed. A decision tree was rendered using the rpart (Recursive Partitioning and Regression Trees) package. Four thousand one hundred twenty-five patients were screened for eligibility. Seventy-three myocarditis patients and 292 nonmyocarditis patients were included. Compared with the control group, the study group was found to have a higher mean respiratory rate (37 ± 23 vs 23 ± 7 breaths per minute) and mean heart rate (121 ± 44 vs 97 ± 25 beats per minute) and lower mean systolic and diastolic blood pressure (102 ± 27/56 ± 17 mm Hg vs 114 ± 14/67 ± 10 mm Hg). The mean white blood cell count was greater in the case group (13 ± 6 vs 10 ± 5 × 10 3 /μL). A decision tree was rendered using simple demographic and clinical variables. The accuracy of the algorithm was 85.2%, with 100% accuracy in patients aged 0 to 2.5 years and 69% in patients aged 2.5 to 18 years. The clinical and laboratory characteristics described in this study were similar to what is described in the literature. The decision tree may aid in the diagnosis of myocarditis in patients 2.5 years and younger. In the population aged 2.5 to 18 years, the decision tree did not constitute an adequate tool for detecting myocarditis.

SANe: Space adaptation network for temporal knowledge graph completion

Temporal Knowledge Graphs (TKGs) model time-dependent facts as relations between entities at specific timestamps, making them well-suited for real-world scenarios. However, TKGs are susceptible to incompleteness, necessitating Temporal Knowledge Graph Completion (TKGC) to predict missing facts. Prior methods often struggle to effectively handle two critical properties of TKGs, time-variability and time-stability, simultaneously, which hinders their performance. In this paper, we propose Space Adaptation Network (SANe), a novel approach for TKGC. SANe adapts facts at different timestamps to distinct latent spaces, effectively addressing time-variability. Our model introduces Parameter Generation Network to produce separate neural networks for each snapshot, which are then encoded into different latent spaces. A dynamic convolutional neural network processes entities and relations, utilizing different learned parameters generated by parameter generation network with respect to timestamps. By handling different temporal snapshots separately, TKGC is transformed into static KGC, enabling the modeling of time-variability. Dynamic convolutional neural network efficiently learns collective knowledge over large periods and supplements more specific knowledge gradually in smaller periods, facilitating time-stability. To strike a balance between learning time-variability and time-stability, we introduce a time-aware parameter generator to produce parameters hierarchically based on year, month, and day timestamps. Long-term knowledge is effectively shared across adjacent snapshots within the same year or month, while short-term knowledge within a day is preserved in specific parameters. However, in unbalanced TKGs, where many facts occur in small intervals, the large number of parameters generated by time-aware parameter generator may remain underutilized. To address this, we propose Adaptive Parameter Generation with a partition tree, ensuring parameter load balancing while maintaining time-stability. We conduct extensive experiments on five benchmark datasets, demonstrating the superiority of SANe over existing methods for TKGC, achieving state-of-the-art performance. Our contributions include pioneering TKGC from the perspective of space adaptation, achieving a balance between time-variability and time-stability through latent space overlap constraints, and substantiating the effectiveness of our model through comprehensive experiments on rich temporal datasets.